sync(hf): snapshot origin main after resource audit cycle

.github/workflows/resource_sync.yml

@@ -4,11 +4,29 @@ on:
   schedule:
     - cron: "0 4 * * *"
   workflow_dispatch:
+    inputs:
+      limit:
+        description: "Candidate fetch limit per source"
+        required: false
+        default: "20"
+      max_promotions:
+        description: "Optional max number of candidate promotions"
+        required: false
+        default: ""
+      enforce_pashto_relevance:
+        description: "Also remove existing entries without Pashto evidence"
+        required: false
+        default: "true"
 
 permissions:
   contents: write
   pull-requests: write
 
+env:
+  RESOURCE_LIMIT: ${{ github.event.inputs.limit || '20' }}
+  MAX_PROMOTIONS: ${{ github.event.inputs.max_promotions || '' }}
+  ENFORCE_PASHTO_RELEVANCE: ${{ github.event.inputs.enforce_pashto_relevance || 'true' }}
+
 jobs:
   sync:
     runs-on: ubuntu-latest
@@ -26,11 +44,24 @@ jobs:
           python -m pip install --upgrade pip
           python -m pip install -e ".[dev]"
 
+      - name: Review existing resources for stale or low-value entries
+        run: |
+          if [ "${ENFORCE_PASHTO_RELEVANCE}" = "true" ]; then
+            python scripts/review_existing_resources.py --enforce-pashto-relevance
+          else
+            python scripts/review_existing_resources.py
+          fi
+
       - name: Sync candidate resources
-        run: python scripts/sync_resources.py --limit 20
+        run: python scripts/sync_resources.py --limit "${RESOURCE_LIMIT}"
 
       - name: Auto-promote valid candidates
-        run: python scripts/promote_candidates.py
+        run: |
+          if [ -n "${MAX_PROMOTIONS}" ]; then
+            python scripts/promote_candidates.py --max-promotions "${MAX_PROMOTIONS}"
+          else
+            python scripts/promote_candidates.py
+          fi
 
       - name: Validate catalog
         run: python scripts/validate_resource_catalog.py
@@ -79,6 +110,7 @@ jobs:
             Automated daily resource sync.
 
             Scope:
+            - Reviews existing catalog entries and removes stale ones only with strong logged reasons
             - Updates `resources/catalog/pending_candidates.json`
             - Auto-promotes valid non-duplicate candidates into `resources/catalog/resources.json`
             - Regenerates resource indexes and search payload
@@ -88,6 +120,7 @@ jobs:
           add-paths: |
             resources/catalog/pending_candidates.json
             resources/catalog/resources.json
+            resources/catalog/removal_log.json
             resources/README.md
             resources/datasets/README.md
             resources/models/README.md

docs/resource_automation.md

@@ -28,6 +28,7 @@ This repository uses automated discovery and promotion to keep Pashto resources
 ## Scripts
 - Validate catalog: `python scripts/validate_resource_catalog.py`
 - Generate markdown and search index: `python scripts/generate_resource_views.py`
+- Review existing resources for stale/deleted entries: `python scripts/review_existing_resources.py`
 - Sync new candidates: `python scripts/sync_resources.py --limit 20`
 - Auto-promote valid candidates: `python scripts/promote_candidates.py`
 - Full run wrapper: `python scripts/run_resource_cycle.py --limit 25`
@@ -39,14 +40,16 @@ This repository uses automated discovery and promotion to keep Pashto resources
   - markdown link checks
   - tests
 - Resource Sync (`.github/workflows/resource_sync.yml`) runs daily, syncs candidates, auto-promotes valid non-duplicate entries, regenerates views, and opens a PR.
+  - Before candidate sync, it reviews existing entries and removes stale/deleted or non-Pashto/low-value entries only with strong logged reasons.
 
 ## Promotion flow
-1. Sync candidates into `resources/catalog/pending_candidates.json`.
-2. Auto-promote valid, non-duplicate entries into `resources/catalog/resources.json`.
-3. Run:
+1. Review existing resources and remove stale entries with strong reasons.
+2. Sync candidates into `resources/catalog/pending_candidates.json`.
+3. Auto-promote valid, non-duplicate, URL-available entries into `resources/catalog/resources.json`.
+4. Run:
    - `python scripts/validate_resource_catalog.py`
    - `python scripts/generate_resource_views.py`
-4. Review PR and merge.
+5. Review PR and merge.
 
 ## Runbook
 - Reusable process guide: [resource_cycle_runbook.md](resource_cycle_runbook.md)

docs/resource_cycle_runbook.md

@@ -5,7 +5,7 @@ Use this runbook whenever you want to repeat the resource update process without
 ## Daily automation (already enabled)
 - Workflow: [../.github/workflows/resource_sync.yml](../.github/workflows/resource_sync.yml)
 - Schedule: every day at 04:00 UTC via GitHub Actions cron.
-- Output: updates [../resources/catalog/pending_candidates.json](../resources/catalog/pending_candidates.json), auto-promotes valid non-duplicate entries into [../resources/catalog/resources.json](../resources/catalog/resources.json), regenerates views, and opens a review PR.
+- Output: reviews existing resources for stale/deleted links and non-Pashto/low-value entries (removing only with strong logged reasons), updates [../resources/catalog/pending_candidates.json](../resources/catalog/pending_candidates.json), auto-promotes valid non-duplicate entries into [../resources/catalog/resources.json](../resources/catalog/resources.json), regenerates views, and opens a review PR.
 
 ## Manual run (single command)
 Run from repository root:
@@ -15,12 +15,13 @@ python scripts/run_resource_cycle.py --limit 25
 ```
 
 What it executes:
-1. `python scripts/sync_resources.py --limit 25`
-2. `python scripts/promote_candidates.py`
-3. `python scripts/validate_resource_catalog.py`
-4. `python scripts/generate_resource_views.py`
-5. `python scripts/check_links.py`
-6. `python -m pytest -q`
+1. `python scripts/review_existing_resources.py`
+2. `python scripts/sync_resources.py --limit 25`
+3. `python scripts/promote_candidates.py`
+4. `python scripts/validate_resource_catalog.py`
+5. `python scripts/generate_resource_views.py`
+6. `python scripts/check_links.py`
+7. `python -m pytest -q`
 
 Candidate sources in the sync step include Kaggle datasets, Hugging Face datasets/models/spaces, GitHub repositories, GitLab repositories, Zenodo records, Dataverse datasets, DataCite DOI records, and paper endpoints (arXiv, Semantic Scholar, OpenAlex, Crossref).
 
@@ -33,7 +34,8 @@ If you want fresh candidates without auto-promotion:
 5. Commit and push.
 
 ## Guardrails
-- Auto-promotion accepts only entries that pass dedupe and catalog validation checks.
+- Auto-promotion accepts only entries that pass dedupe, URL-availability checks, and catalog validation checks.
+- Existing resources are auto-removed only for strong reasons (for example confirmed hard-missing links, duplicates, or missing Pashto relevance), with reasons stored in `resources/catalog/removal_log.json`.
 - Keep `status: verified` for entries that pass automation checks and repository review.
 - Do not promote "reference-only" resources where Pashto is incidental; only Pashto-centric resources are eligible.
 - Treat spelling variants as valid Pashto markers during review (`pashto`, `pukhto`, `pushto`, `pakhto`, `pashto-script`).

resources/README.md

@@ -3,13 +3,13 @@
 Structured, Pashto-focused resource tracking lives in this folder.
 
 ## Sections
-- Datasets (48): [datasets/README.md](datasets/README.md)
-- Models (18): [models/README.md](models/README.md)
+- Datasets (46): [datasets/README.md](datasets/README.md)
+- Models (19): [models/README.md](models/README.md)
 - Benchmarks (4): [benchmarks/README.md](benchmarks/README.md)
 - Tools (0): [tools/README.md](tools/README.md)
-- Papers (24): [papers/README.md](papers/README.md)
-- Projects (17): [projects/README.md](projects/README.md)
-- Code (1): [codes/README.md](codes/README.md)
+- Papers (104): [papers/README.md](papers/README.md)
+- Projects (48): [projects/README.md](projects/README.md)
+- Code (4): [codes/README.md](codes/README.md)
 
 ## Machine-Readable Catalog
 - Canonical catalog: [catalog/resources.json](catalog/resources.json)
@@ -22,4 +22,4 @@ Structured, Pashto-focused resource tracking lives in this folder.
 - Run `python scripts/validate_resource_catalog.py` before opening a PR.
 - Run `python scripts/generate_resource_views.py` after catalog changes.
 
-Verified resource count: `112`
+Verified resource count: `225`

resources/catalog/removal_log.json

@@ -0,0 +1,1205 @@
+{
+  "updated_on": "2026-02-22",
+  "entries": [
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-pashto-isolated-words",
+      "title": "Pashto Isolated Words Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-pashto-word-embeddings",
+      "title": "Pashto Word Embeddings",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-pold-pashto-offensive",
+      "title": "POLD - Pashto Offensive Language Dataset",
+      "url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-pashto-english-sentiment-corpus",
+      "title": "Pashto English Bilingual Sentiment Corpus",
+      "url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-urdu-pashto-lexicon",
+      "title": "Urdu-Pashto Lexicon Dataset",
+      "url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-drijaz-pashtoocr",
+      "title": "PashtoOCR (Kaggle)",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-english-pashto-language-dataset-epld",
+      "title": "English-Pashto Language Dataset (EPLD)",
+      "url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-katib-s-pashto-text-imagebase-kpti",
+      "title": "Katib's Pashto Text Imagebase (KPTI)",
+      "url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-pashto-ocr",
+      "title": "Pashto OCR",
+      "url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "dataset-kaggle-common-voice-24-0-pashto-speech-dataset",
+      "title": "Common Voice 24.0: Pashto Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "candidate-kaggle-dataset-abdulbasitkh-pashto-isolated-alphabets-and-numerals",
+      "title": "Pashto Isolated Alphabets and Numerals",
+      "url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "candidate-kaggle-dataset-alimuhammadasad-pashto-poetry",
+      "title": "Pashto Poetry",
+      "url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "candidate-kaggle-dataset-mahibullahmudaser-pashto-text-characters-sample",
+      "title": "Pashto text characters sample",
+      "url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "candidate-kaggle-dataset-ahmadferozafshar-pashto-language-alphabets",
+      "title": "pashto_language_alphabets",
+      "url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:47:22.435531+00:00",
+      "id": "candidate-kaggle-dataset-aimalrezvan-pashto-language-characters",
+      "title": "Pashto_language_characters",
+      "url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-ataullahaali-common-voice-24-0-pashto-speech-dataset",
+      "title": "Common Voice 24.0: Pashto Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-rabiakhan827-english-pashto-language-dataset-epld",
+      "title": "English-Pashto Language Dataset (EPLD)",
+      "url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-hassanamin-katib-s-pashto-text-imagebase-kpti",
+      "title": "Katib's Pashto Text Imagebase (KPTI)",
+      "url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-farhadkhan66-pashto-english-bilingual-sentiment-corpus",
+      "title": "Pashto English Bilingual Sentiment Corpus",
+      "url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-abdulbasitkh-pashto-isolated-alphabets-and-numerals",
+      "title": "Pashto Isolated Alphabets and Numerals",
+      "url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-engrirf-pashto-isolated-words-speech-dataset",
+      "title": "Pashto Isolated Words Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-hassanamin-pashto-ocr",
+      "title": "Pashto OCR",
+      "url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-alimuhammadasad-pashto-poetry",
+      "title": "Pashto Poetry",
+      "url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-mahibullahmudaser-pashto-text-characters-sample",
+      "title": "Pashto text characters sample",
+      "url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pashto-word-embeddings",
+      "title": "Pashto Word Embeddings",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-ahmadferozafshar-pashto-language-alphabets",
+      "title": "pashto_language_alphabets",
+      "url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-aimalrezvan-pashto-language-characters",
+      "title": "Pashto_language_characters",
+      "url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pashtoocr",
+      "title": "PashtoOCR",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pold-pashto-offensive-language-dataset",
+      "title": "POLD - Pashto Offensive Language Dataset",
+      "url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T19:59:50.593781+00:00",
+      "id": "candidate-kaggle-dataset-shafeeqgigyani-urdu-pashto-lexicon-dataset",
+      "title": "Urdu-Pashto Lexicon Dataset",
+      "url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-ataullahaali-common-voice-24-0-pashto-speech-dataset",
+      "title": "Common Voice 24.0: Pashto Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-rabiakhan827-english-pashto-language-dataset-epld",
+      "title": "English-Pashto Language Dataset (EPLD)",
+      "url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-hassanamin-katib-s-pashto-text-imagebase-kpti",
+      "title": "Katib's Pashto Text Imagebase (KPTI)",
+      "url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-farhadkhan66-pashto-english-bilingual-sentiment-corpus",
+      "title": "Pashto English Bilingual Sentiment Corpus",
+      "url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-abdulbasitkh-pashto-isolated-alphabets-and-numerals",
+      "title": "Pashto Isolated Alphabets and Numerals",
+      "url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-engrirf-pashto-isolated-words-speech-dataset",
+      "title": "Pashto Isolated Words Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-hassanamin-pashto-ocr",
+      "title": "Pashto OCR",
+      "url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-alimuhammadasad-pashto-poetry",
+      "title": "Pashto Poetry",
+      "url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-mahibullahmudaser-pashto-text-characters-sample",
+      "title": "Pashto text characters sample",
+      "url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pashto-word-embeddings",
+      "title": "Pashto Word Embeddings",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-ahmadferozafshar-pashto-language-alphabets",
+      "title": "pashto_language_alphabets",
+      "url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-aimalrezvan-pashto-language-characters",
+      "title": "Pashto_language_characters",
+      "url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pashtoocr",
+      "title": "PashtoOCR",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pold-pashto-offensive-language-dataset",
+      "title": "POLD - Pashto Offensive Language Dataset",
+      "url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:13:47.457104+00:00",
+      "id": "candidate-kaggle-dataset-shafeeqgigyani-urdu-pashto-lexicon-dataset",
+      "title": "Urdu-Pashto Lexicon Dataset",
+      "url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-ataullahaali-common-voice-24-0-pashto-speech-dataset",
+      "title": "Common Voice 24.0: Pashto Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-rabiakhan827-english-pashto-language-dataset-epld",
+      "title": "English-Pashto Language Dataset (EPLD)",
+      "url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-hassanamin-katib-s-pashto-text-imagebase-kpti",
+      "title": "Katib's Pashto Text Imagebase (KPTI)",
+      "url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-farhadkhan66-pashto-english-bilingual-sentiment-corpus",
+      "title": "Pashto English Bilingual Sentiment Corpus",
+      "url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-abdulbasitkh-pashto-isolated-alphabets-and-numerals",
+      "title": "Pashto Isolated Alphabets and Numerals",
+      "url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-engrirf-pashto-isolated-words-speech-dataset",
+      "title": "Pashto Isolated Words Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-hassanamin-pashto-ocr",
+      "title": "Pashto OCR",
+      "url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-alimuhammadasad-pashto-poetry",
+      "title": "Pashto Poetry",
+      "url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-mahibullahmudaser-pashto-text-characters-sample",
+      "title": "Pashto text characters sample",
+      "url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pashto-word-embeddings",
+      "title": "Pashto Word Embeddings",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-ahmadferozafshar-pashto-language-alphabets",
+      "title": "pashto_language_alphabets",
+      "url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-aimalrezvan-pashto-language-characters",
+      "title": "Pashto_language_characters",
+      "url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pashtoocr",
+      "title": "PashtoOCR",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-drijaz-pold-pashto-offensive-language-dataset",
+      "title": "POLD - Pashto Offensive Language Dataset",
+      "url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:27:10.672699+00:00",
+      "id": "candidate-kaggle-dataset-shafeeqgigyani-urdu-pashto-lexicon-dataset",
+      "title": "Urdu-Pashto Lexicon Dataset",
+      "url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-pashto-isolated-words",
+      "title": "Pashto Isolated Words Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-pashto-word-embeddings",
+      "title": "Pashto Word Embeddings",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-pold-pashto-offensive",
+      "title": "POLD - Pashto Offensive Language Dataset",
+      "url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-pashto-english-sentiment-corpus",
+      "title": "Pashto English Bilingual Sentiment Corpus",
+      "url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-urdu-pashto-lexicon",
+      "title": "Urdu-Pashto Lexicon Dataset",
+      "url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-drijaz-pashtoocr",
+      "title": "PashtoOCR (Kaggle)",
+      "url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/drijaz/pashtoocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-english-pashto-language-dataset-epld",
+      "title": "English-Pashto Language Dataset (EPLD)",
+      "url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-katib-s-pashto-text-imagebase-kpti",
+      "title": "Katib's Pashto Text Imagebase (KPTI)",
+      "url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-pashto-ocr",
+      "title": "Pashto OCR",
+      "url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/hassanamin/pashto-ocr",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "dataset-kaggle-common-voice-24-0-pashto-speech-dataset",
+      "title": "Common Voice 24.0: Pashto Speech Dataset",
+      "url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "candidate-kaggle-dataset-abdulbasitkh-pashto-isolated-alphabets-and-numerals",
+      "title": "Pashto Isolated Alphabets and Numerals",
+      "url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "candidate-kaggle-dataset-alimuhammadasad-pashto-poetry",
+      "title": "Pashto Poetry",
+      "url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "candidate-kaggle-dataset-mahibullahmudaser-pashto-text-characters-sample",
+      "title": "Pashto text characters sample",
+      "url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "candidate-kaggle-dataset-ahmadferozafshar-pashto-language-alphabets",
+      "title": "pashto_language_alphabets",
+      "url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    },
+    {
+      "removed_on": "2026-02-21T20:47:45.952635+00:00",
+      "id": "candidate-kaggle-dataset-aimalrezvan-pashto-language-characters",
+      "title": "Pashto_language_characters",
+      "url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+      "reasons": [
+        "URL returned hard-missing HTTP status 404."
+      ],
+      "evidence": {
+        "status_code": 404,
+        "final_url": "https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters",
+        "metadata_pashto": true,
+        "direct_pashto": true,
+        "page_pashto": false
+      }
+    }
+  ]
+}

resources/codes/README.md

@@ -4,7 +4,10 @@
 
 | Resource | Link | Pashto Evidence | Primary Use |
 |---|---|---|---|
+| LGUG2Z/tashkil | [github](https://github.com/LGUG2Z/tashkil) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/LGUG2Z/tashkil) | Automated discovery entry for Pashto resource tracking. |
+| mrychlik/worldly-ocr | [github](https://github.com/mrychlik/worldly-ocr) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/mrychlik/worldly-ocr) | Automated discovery entry for Pashto resource tracking. |
 | nlpashto Toolkit | [github](https://github.com/ijazul-haq/nlpashto) | [Repository name and description explicitly identify a Pashto NLP toolkit. (`Pashto`, `NLP`)](https://api.github.com/repos/ijazul-haq/nlpashto) | Pashto NLP code integration and experimentation |
+| sinaahmadi/PersoArabicLID | [github](https://github.com/sinaahmadi/PersoArabicLID) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/sinaahmadi/PersoArabicLID) | Automated discovery entry for Pashto resource tracking. |
 
 ## Maintenance
 - Source of truth: [../catalog/resources.json](../catalog/resources.json)

resources/datasets/README.md

@@ -14,10 +14,9 @@
 | arsalagrey/pashto-books | [huggingface](https://huggingface.co/datasets/arsalagrey/pashto-books) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/arsalagrey/pashto-books) | Automated discovery entry for Pashto resource tracking. |
 | arsalagrey/pashto-books-json | [huggingface](https://huggingface.co/datasets/arsalagrey/pashto-books-json) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/arsalagrey/pashto-books-json) | Automated discovery entry for Pashto resource tracking. |
 | Belebele | [huggingface](https://huggingface.co/datasets/facebook/belebele) | [Dataset includes pbt_Arab subset. (`pbt_Arab`)](https://huggingface.co/datasets/facebook/belebele) | Comprehension and multilingual NLP benchmark |
-| Common Voice 24.0: Pashto Speech Dataset | [kaggle](https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/ataullahaali/common-voice-scripted-speech-24-0-pashto) | ASR training and evaluation data source |
+| Clitic Particles and the Typology of 2P Languages | [zenodo](https://zenodo.org/records/15010591) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/15010591) | Automated discovery entry for Pashto resource tracking. |
 | Common Voice Scripted Speech 24.0 - Pashto | [mozilla](https://datacollective.mozillafoundation.org/datasets/cmj8u3pnb00llnxxbfvxo3b14) | [Official dataset page is for Pashto. (`Pashto`)](https://datacollective.mozillafoundation.org/datasets/cmj8u3pnb00llnxxbfvxo3b14) | ASR training and evaluation |
 | English to Pashto Sentences Dataset | [huggingface](https://huggingface.co/datasets/adnankhan769/english_to_pashto_sentences_dataset) | [Dataset ID explicitly states English-to-Pashto and includes Pashto-script sentence column. (`Pashto`)](https://huggingface.co/api/datasets/adnankhan769/english_to_pashto_sentences_dataset) | MT and bilingual sentence alignment baseline |
-| English-Pashto Language Dataset (EPLD) | [kaggle](https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/rabiakhan827/english-pashto-language-dataset-epld) | Machine translation and bilingual corpus development |
 | Google FLEURS | [huggingface](https://huggingface.co/datasets/google/fleurs) | [Dataset config includes ps_af. (`ps_af`)](https://huggingface.co/datasets/google/fleurs/blob/main/fleurs.py) | Speech benchmark and external evaluation |
 | IARPA Babel Pashto Language Pack IARPA-babel104b-v0.4bY | [dataverse](https://hdl.handle.net/11272.1/AB2/GLFN3X) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://hdl.handle.net/11272.1/AB2/GLFN3X) | Pashto speech dataset for ASR and language identification experiments |
 | ihanif/pashto_asr_wer | [huggingface](https://huggingface.co/datasets/ihanif/pashto_asr_wer) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/ihanif/pashto_asr_wer) | ASR training and evaluation data source |
@@ -27,29 +26,28 @@
 | ihanif/pashto_speech_5k | [huggingface](https://huggingface.co/datasets/ihanif/pashto_speech_5k) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/ihanif/pashto_speech_5k) | ASR training and evaluation data source |
 | ihanif/pashto_speech_ds | [huggingface](https://huggingface.co/datasets/ihanif/pashto_speech_ds) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/ihanif/pashto_speech_ds) | ASR training and evaluation data source |
 | ihanif/pashto_speech_parquet_10k | [huggingface](https://huggingface.co/datasets/ihanif/pashto_speech_parquet_10k) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/ihanif/pashto_speech_parquet_10k) | ASR training and evaluation data source |
-| Katib's Pashto Text Imagebase (KPTI) | [kaggle](https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/hassanamin/katibs-pashto-text-imagebase-kpti) | OCR training and evaluation data source |
 | koochikoo25/Pashto-Concatenated | [huggingface](https://huggingface.co/datasets/koochikoo25/Pashto-Concatenated) | [Dataset title explicitly states Pashto and card metadata exposes audio-text features and splits. (`Pashto`, `audio`, `transcription`)](https://huggingface.co/datasets/koochikoo25/Pashto-Concatenated) | ASR dataset preparation and split-based benchmark experiments |
+| Multi-Language Conversational Telephone Speech 2011 -- Central Asian | [dataverse](https://hdl.handle.net/11272.1/AB2/YW9PX3) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://hdl.handle.net/11272.1/AB2/YW9PX3) | Automated discovery entry for Pashto resource tracking. |
+| NAVOIY-TERRA Corpus v1.0: First Computational Corpus of Alisher Navoi Works with Nine-Language Semantic Annotations | [datacite](https://zenodo.org/doi/10.5281/zenodo.18602634) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/doi/10.5281/zenodo.18602634) | Automated discovery entry for Pashto resource tracking. |
 | oowais/pushto-text-to-speech-dataset | [huggingface](https://huggingface.co/datasets/oowais/pushto-text-to-speech-dataset) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/oowais/pushto-text-to-speech-dataset) | ASR training and evaluation data source |
 | OPED (Open Pashto-English Dictionary): Preliminary version, 30 October 2025 | [zenodo](https://zenodo.org/records/17487678) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/17487678) | Automated discovery entry for Pashto resource tracking. |
 | OPUS-100 | [huggingface](https://huggingface.co/datasets/Helsinki-NLP/opus-100) | [Dataset viewer includes en-ps split. (`en-ps`)](https://huggingface.co/datasets/Helsinki-NLP/opus-100/viewer/en-ps) | Machine translation training and evaluation |
 | OSCAR Corpus | [huggingface](https://huggingface.co/datasets/oscar-corpus/oscar) | [Dataset includes unshuffled_deduplicated_ps split. (`unshuffled_deduplicated_ps`)](https://huggingface.co/datasets/oscar-corpus/oscar) | Language modeling and lexicon expansion |
-| Pashto English Bilingual Sentiment Corpus | [kaggle](https://www.kaggle.com/datasets/farhadkhan66/pashto-translated-corpus) | [Kaggle dataset title and description identify the corpus as Pashto-English sentiment data. (`Pashto`)](https://www.kaggle.com/api/v1/datasets/view/farhadkhan66/pashto-translated-corpus) | Sentiment analysis and bilingual NLP experiments |
-| Pashto Isolated Alphabets and Numerals | [kaggle](https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/abdulbasitkh/pashto-isolated-alphabetss-and-numerals) | Automated discovery entry for Pashto resource tracking. |
-| Pashto Isolated Words Speech Dataset | [kaggle](https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset) | [Dataset title explicitly states Pashto speech dataset. (`Pashto`)](https://www.kaggle.com/datasets/engrirf/pashto-isolated-words-speech-dataset) | Keyword spotting and constrained ASR experiments |
-| Pashto OCR | [kaggle](https://www.kaggle.com/datasets/hassanamin/pashto-ocr) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/hassanamin/pashto-ocr) | OCR training and evaluation data source |
-| Pashto Poetry | [kaggle](https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/alimuhammadasad/pashto-poetry) | Automated discovery entry for Pashto resource tracking. |
-| Pashto text characters sample | [kaggle](https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/mahibullahmudaser/pashtochracterssample) | Automated discovery entry for Pashto resource tracking. |
 | Pashto Wikipedia Corpus | [huggingface](https://huggingface.co/datasets/ihanif/pashto-wikipedia-corpus) | [Dataset metadata includes language:ps and the title specifies Pashto corpus. (`ps`, `Pashto`)](https://huggingface.co/datasets/ihanif/pashto-wikipedia-corpus) | Pashto text corpus for NLP baselines |
-| Pashto Word Embeddings | [kaggle](https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings) | [Dataset description states pretrained Pashto embeddings. (`Pashto`)](https://www.kaggle.com/datasets/drijaz/pashto-word-embeddings) | Lexical semantics and lightweight NLP baselines |
-| pashto_language_alphabets | [kaggle](https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/ahmadferozafshar/pashto-language-alphabets) | Automated discovery entry for Pashto resource tracking. |
-| Pashto_language_characters | [kaggle](https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters) | [Kaggle dataset title/subtitle includes Pashto keyword. (`Pashto`)](https://www.kaggle.com/datasets/aimalrezvan/pashto-language-characters) | Automated discovery entry for Pashto resource tracking. |
-| PashtoOCR (Kaggle) | [kaggle](https://www.kaggle.com/datasets/drijaz/pashtoocr) | [Kaggle dataset title and subtitle explicitly identify a Pashto OCR dataset. (`Pashto`, `OCR`)](https://www.kaggle.com/api/v1/datasets/view/drijaz/pashtoocr) | Pashto OCR dataset benchmarking and training |
-| POLD - Pashto Offensive Language Dataset | [kaggle](https://www.kaggle.com/datasets/drijaz/pold-pashto-offensive-language-dataset) | [Kaggle title and description explicitly state Pashto offensive language benchmark dataset. (`Pashto`)](https://www.kaggle.com/api/v1/datasets/view/drijaz/pold-pashto-offensive-language-dataset) | Pashto toxicity and moderation NLP benchmarks |
+| RATS Language Identification | [dataverse](https://hdl.handle.net/11272.1/AB2/UP3WJC) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://hdl.handle.net/11272.1/AB2/UP3WJC) | Automated discovery entry for Pashto resource tracking. |
+| RATS Low Speech Density | [dataverse](https://doi.org/10.35111/4ENA-FG30) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://doi.org/10.35111/4ENA-FG30) | Automated discovery entry for Pashto resource tracking. |
+| RATS Speaker Identification | [dataverse](https://doi.org/10.35111/ZQET-2102) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://doi.org/10.35111/ZQET-2102) | Automated discovery entry for Pashto resource tracking. |
+| RATS Speech Activity Detection | [dataverse](https://hdl.handle.net/11272.1/AB2/1UISJ7) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://hdl.handle.net/11272.1/AB2/1UISJ7) | Automated discovery entry for Pashto resource tracking. |
 | saillab/alpaca_pashto_taco | [huggingface](https://huggingface.co/datasets/saillab/alpaca_pashto_taco) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/saillab/alpaca_pashto_taco) | Instruction tuning and LLM adaptation data source |
 | SherwinDesouza/pashto-common-voice-20 | [huggingface](https://huggingface.co/datasets/SherwinDesouza/pashto-common-voice-20) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/SherwinDesouza/pashto-common-voice-20) | Pashto data source for NLP experimentation |
+| SMAjram: A Large-Scale Synthetic OCR Dataset for Punjabi Shahmukhi (Perso-Arabic) Script | [zenodo](https://zenodo.org/records/15868719) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/15868719) | Automated discovery entry for Pashto resource tracking. |
 | tasal9/Pashto_Dataset | [huggingface](https://huggingface.co/datasets/tasal9/Pashto_Dataset) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/tasal9/Pashto_Dataset) | Pashto data source for NLP experimentation |
 | tasal9/ZamAI_Pashto_Dataset | [huggingface](https://huggingface.co/datasets/tasal9/ZamAI_Pashto_Dataset) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/datasets/tasal9/ZamAI_Pashto_Dataset) | Pashto data source for NLP experimentation |
-| Urdu-Pashto Lexicon Dataset | [kaggle](https://www.kaggle.com/datasets/shafeeqgigyani/urdu-pashto-lexicon-dataset) | [Kaggle metadata describes 7,601 Urdu entries with Pashto translations. (`Pashto`)](https://www.kaggle.com/api/v1/datasets/view/shafeeqgigyani/urdu-pashto-lexicon-dataset) | Lexicon and translation lexeme mapping |
+| Towards a Typology of Endoclitics | [zenodo](https://zenodo.org/records/15041544) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/15041544) | Automated discovery entry for Pashto resource tracking. |
+| TRAD Arabic-French Parallel Text -- Newsgroup | [dataverse](https://hdl.handle.net/11272.1/AB2/0DET8M) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://hdl.handle.net/11272.1/AB2/0DET8M) | Automated discovery entry for Pashto resource tracking. |
+| TRAD Arabic-French Parallel Text -- Newswire | [dataverse](https://doi.org/10.35111/Z1WG-9X78) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://doi.org/10.35111/Z1WG-9X78) | Automated discovery entry for Pashto resource tracking. |
+| TRAD Chinese-French Parallel Text -- Blog | [dataverse](https://hdl.handle.net/11272.1/AB2/ATYE6I) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://hdl.handle.net/11272.1/AB2/ATYE6I) | Automated discovery entry for Pashto resource tracking. |
+| TRAD Chinese-French Parallel Text -- Broadcast News | [dataverse](https://doi.org/10.35111/7FW4-EV85) | [Dataverse metadata includes Pashto markers in dataset title or description. (`pashto`)](https://doi.org/10.35111/7FW4-EV85) | Automated discovery entry for Pashto resource tracking. |
 | Wikimedia Wikipedia | [huggingface](https://huggingface.co/datasets/wikimedia/wikipedia) | [Dataset includes 20231101.ps subset. (`20231101.ps`)](https://huggingface.co/datasets/wikimedia/wikipedia) | Terminology and balanced text corpus |
 | Zirak-AI PashtoOCR | [huggingface](https://huggingface.co/datasets/zirak-ai/PashtoOCR) | [Dataset tags include language:ps and the dataset name is PashtoOCR. (`ps`, `PashtoOCR`)](https://huggingface.co/datasets/zirak-ai/PashtoOCR) | OCR and text extraction benchmarking |
 

resources/models/README.md

@@ -13,6 +13,7 @@
 | ihanif/whisper-small-pashto-dropout | [huggingface](https://huggingface.co/ihanif/whisper-small-pashto-dropout) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ihanif/whisper-small-pashto-dropout) | Pashto ASR baseline and model comparison |
 | ihanif/xls-r-1b-pashto | [huggingface](https://huggingface.co/ihanif/xls-r-1b-pashto) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ihanif/xls-r-1b-pashto) | Pashto ASR baseline and model comparison |
 | ijazulhaq/bert-base-pashto | [huggingface](https://huggingface.co/ijazulhaq/bert-base-pashto) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ijazulhaq/bert-base-pashto) | Pashto model baseline for downstream NLP tasks |
+| ijazulhaq/bert-base-pashto-c | [huggingface](https://huggingface.co/ijazulhaq/bert-base-pashto-c) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ijazulhaq/bert-base-pashto-c) | Automated discovery entry for Pashto resource tracking. |
 | ijazulhaq/bert-base-pashto-v1 | [huggingface](https://huggingface.co/ijazulhaq/bert-base-pashto-v1) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/ijazulhaq/bert-base-pashto-v1) | Pashto model baseline for downstream NLP tasks |
 | Jawaria/wav2vec2-large-xls-r-300m-pashto-colab-final-1 | [huggingface](https://huggingface.co/Jawaria/wav2vec2-large-xls-r-300m-pashto-colab-final-1) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/Jawaria/wav2vec2-large-xls-r-300m-pashto-colab-final-1) | Automated discovery entry for Pashto resource tracking. |
 | koochikoo25/pashto-whisper-large | [huggingface](https://huggingface.co/koochikoo25/pashto-whisper-large) | [Matched by Pashto keyword in Hugging Face search results. (`pashto`)](https://huggingface.co/koochikoo25/pashto-whisper-large) | Pashto ASR baseline and model comparison |

resources/papers/README.md

@@ -4,30 +4,110 @@
 
 | Resource | Link | Pashto Evidence | Primary Use |
 |---|---|---|---|
+| (Pushto) Pakhto Nasar Kay Da Matbooa Tarjumo Yova Tanqeedi Mutala/Jaiza. | [other](https://www.semanticscholar.org/paper/0da0e8535262d1f26f04dd6bc2f091474cab4150) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/0da0e8535262d1f26f04dd6bc2f091474cab4150) | Automated discovery entry for Pashto resource tracking. |
+| A Comparative Analysis of Pashto Ghazals and English Sonnets in 17th Century | [other](https://www.semanticscholar.org/paper/55b044485b2f134c69c9b9b6dfeaa7e71e704b3d) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/55b044485b2f134c69c9b9b6dfeaa7e71e704b3d) | Automated discovery entry for Pashto resource tracking. |
+| A Dictionary of the Pukhto, Pushto, or Language of the Afghans | [other](https://www.semanticscholar.org/paper/777c0aa56991f55826339915363de2ceb8dd7141) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/777c0aa56991f55826339915363de2ceb8dd7141) | Automated discovery entry for Pashto resource tracking. |
+| A dictionary of the Pukhto, Pushto, or language of the Afghans; with remarks on the originality of the language, and its affinity to the Semitic and other Oriental tongues, etc. | [other](https://www.semanticscholar.org/paper/d12502a6c245ff6f537bf68d9db4b449dca827bb) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/d12502a6c245ff6f537bf68d9db4b449dca827bb) | Automated discovery entry for Pashto resource tracking. |
+| A grammar of the Puk̲h̲to or Pus̲'h̲to language | [other](https://www.semanticscholar.org/paper/99c46409a55ac0bf68e2c530a377becfcb46dd47) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/99c46409a55ac0bf68e2c530a377becfcb46dd47) | Automated discovery entry for Pashto resource tracking. |
+| A New Etymological Vocabulary of Pashto | [openalex](https://openalex.org/W2071464713) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://openalex.org/W2071464713) | Automated discovery entry for Pashto resource tracking. |
+| A reference grammar of Pashto | [openalex](http://wals.info/refdb/record/7189) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](http://wals.info/refdb/record/7189) | Automated discovery entry for Pashto resource tracking. |
+| An Acoustic Analysis of consonants of Khattak Dialect of Pashto | [other](https://www.semanticscholar.org/paper/ed06d206e60a62c2bebdd487b4f8dea253a9a0a8) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/ed06d206e60a62c2bebdd487b4f8dea253a9a0a8) | Automated discovery entry for Pashto resource tracking. |
+| AN ANALYSIS OF FREUDIAN CONCEPT OF MOURNING IN PASHTO TAPPAS ON THE THEME OF MIGRATION | [zenodo](https://zenodo.org/records/11124039) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/11124039) | Automated discovery entry for Pashto resource tracking. |
+| An Analysis of the Syntactic and Pragmatic Effects on Word Order Flexibility in Pashto and English | [other](https://www.semanticscholar.org/paper/136c23f176399f7dfc45e6ae990a975aafd7da1d) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/136c23f176399f7dfc45e6ae990a975aafd7da1d) | Automated discovery entry for Pashto resource tracking. |
+| Analysing Deep Meaning of Proverbs in Pashto Language | [other](https://www.semanticscholar.org/paper/1a804a9701c5103ed38df3350da61abdf5df2b57) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/1a804a9701c5103ed38df3350da61abdf5df2b57) | Automated discovery entry for Pashto resource tracking. |
 | Benchmark Pashto Handwritten Character Dataset and Pashto Object Character Recognition (OCR) Using Deep Neural Network with Rule Activation Function | [openalex](https://doi.org/10.1155/2021/6669672) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1155/2021/6669672) | Pashto handwritten OCR benchmark and methodology reference |
 | Benchmarking Whisper for Low-Resource Speech Recognition: An N-Shot Evaluation on Pashto, Punjabi, and Urdu | [other](https://www.semanticscholar.org/paper/13104eddc785756132a19242ac7e74442b145693) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/13104eddc785756132a19242ac7e74442b145693) | Pashto research reference for methods and benchmarking |
+| CER-HV: A CER-Based Human-in-the-Loop Framework for Cleaning Datasets Applied to Arabic-Script HTR | [datacite](https://arxiv.org/abs/2601.16713) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://arxiv.org/abs/2601.16713) | Automated discovery entry for Pashto resource tracking. |
+| CHALLENGING GENDER ROLES: A FEMINIST ANALYSIS OF GHANI KHAN'S THE PATHANS | [zenodo](https://zenodo.org/records/11216862) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/11216862) | Automated discovery entry for Pashto resource tracking. |
+| Cinematic Misnomers: Examining the Effects of Pashto Movie Titles on the Perception of Pashtun Identity | [other](https://www.semanticscholar.org/paper/1b4c38ce4ceb6ac7846062bb589351cc88a36617) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/1b4c38ce4ceb6ac7846062bb589351cc88a36617) | Automated discovery entry for Pashto resource tracking. |
+| Comparative Study of Adjectives in Pashto and Dari as Cognate Languages | [other](https://www.semanticscholar.org/paper/558e9dd7d4027be391a39f5e5ef988cf05039dc7) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/558e9dd7d4027be391a39f5e5ef988cf05039dc7) | Automated discovery entry for Pashto resource tracking. |
+| Comprehensive Socio-phonetic Study of the Plosive /p/ and Fricative /f/ Merger among Pashto Speakers in Khyber Pakhtunkhwa | [other](https://www.semanticscholar.org/paper/4f01f2250c897dc53099f76a2455471b480f22cf) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/4f01f2250c897dc53099f76a2455471b480f22cf) | Automated discovery entry for Pashto resource tracking. |
+| Critical study of the travelogues of Dr Altaf Yousafzai (In The Context of "Thailand kay Rang", "Nile kay Sang" and  "Bakhal-e-Hinduwush Bakhsham") | [zenodo](https://zenodo.org/records/13937101) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/13937101) | Automated discovery entry for Pashto resource tracking. |
 | Database development and automatic speech recognition of isolated Pashto spoken digits using MFCC and K-NN | [openalex](https://doi.org/10.1007/s10772-014-9267-z) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1007/s10772-014-9267-z) | Pashto ASR baseline method reference for digit recognition |
 | Deep Learning-Based Detection of One and Two-Column Textual Blocks in Camera-Captured Pashto Documents Images | [other](https://www.semanticscholar.org/paper/8c9d2628e23d5c27edc656071f11f0e78124d182) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/8c9d2628e23d5c27edc656071f11f0e78124d182) | Pashto research reference for methods and benchmarking |
+| Depiction of Women's Cries in Pashto Landai Poetry | [zenodo](https://zenodo.org/records/15524281) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/15524281) | Automated discovery entry for Pashto resource tracking. |
+| Descriptive Grammar of Pashto and its Dialects | [openalex](https://doi.org/10.1515/9781614512318) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1515/9781614512318) | Automated discovery entry for Pashto resource tracking. |
 | Development of a New Image-to-text Conversion System for Pashto, Farsi and Traditional Chinese | [arxiv](http://arxiv.org/abs/2005.08650v1) | [Matched by Pashto marker in paper title from arXiv query results. (`pashto`)](http://arxiv.org/abs/2005.08650v1) | Pashto OCR method reference |
+| Divorce And Women’s Rights: Should Women have Equal Rights? (Pashto) | [datacite](https://opendata.uni-halle.de//handle/1981185920/123991) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://opendata.uni-halle.de//handle/1981185920/123991) | Automated discovery entry for Pashto resource tracking. |
+| Doing Pashto | [crossref](https://doi.org/10.1080/02690055.2015.1068987) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.1080/02690055.2015.1068987) | Automated discovery entry for Pashto resource tracking. |
+| Editorial Note (Dari) | [datacite](https://opendata.uni-halle.de//handle/1981185920/123994) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://opendata.uni-halle.de//handle/1981185920/123994) | Automated discovery entry for Pashto resource tracking. |
+| Editorial Note (English) | [datacite](https://opendata.uni-halle.de//handle/1981185920/123993) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://opendata.uni-halle.de//handle/1981185920/123993) | Automated discovery entry for Pashto resource tracking. |
+| Editorial Note (Pashto) | [datacite](https://opendata.uni-halle.de//handle/1981185920/123995) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://opendata.uni-halle.de//handle/1981185920/123995) | Automated discovery entry for Pashto resource tracking. |
+| EDUCATIONAL AND LINGUISTIC ASPECTS OF TEXT PREPROCESSING IN PASHTO | [zenodo](https://zenodo.org/records/15917449) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/15917449) | Automated discovery entry for Pashto resource tracking. |
+| Embedding Elements from Foreign Language into The Native Language Through Pashto-English Code-Mixed Speech | [zenodo](https://zenodo.org/records/14756994) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/14756994) | Automated discovery entry for Pashto resource tracking. |
 | Enhancing Pashto NER Using Machine-Labeled Data and Transformer-Based Models | [other](https://www.semanticscholar.org/paper/be851ecf9197ef9bb8bf764abf4db0dda95cd9da) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/be851ecf9197ef9bb8bf764abf4db0dda95cd9da) | Pashto research reference for methods and benchmarking |
 | Enhancing Pashto Text Classification using Language Processing Techniques for Single And Multi-Label Analysis | [arxiv](http://arxiv.org/abs/2305.03201v1) | [Matched by arXiv query: all:pashto. (`pashto`)](http://arxiv.org/abs/2305.03201v1) | Pashto research reference for methods and benchmarking |
+| Essential Skills for a Lexicographer: Based on Pashto Lexicography | [other](https://www.semanticscholar.org/paper/8fc45aa567cb78713e2fef41d5e748e8ee1d8470) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/8fc45aa567cb78713e2fef41d5e748e8ee1d8470) | Automated discovery entry for Pashto resource tracking. |
+| EVALUATION OF ANTIBACTERIAL ACTIVITY OF ZIZYPHUS JUJUBA | [zenodo](https://zenodo.org/records/3595881) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/3595881) | Automated discovery entry for Pashto resource tracking. |
+| EVALUATION OF ANTIPYRETIC ACTIVITY OF ZIZYPHUS JUJUBA LAM. LEAVES ON ALBINO RATS | [zenodo](https://zenodo.org/records/4269214) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/4269214) | Automated discovery entry for Pashto resource tracking. |
+| Exploring Hospitality as a Cultural Tradition: A Qualitative Study of Pashto and Hindko Customs | [zenodo](https://zenodo.org/records/14872725) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/14872725) | Automated discovery entry for Pashto resource tracking. |
+| Exploring the Impacts of Emotion through Language Learning on Pashto Speakers Young Adulthood in District Peshawar | [other](https://www.semanticscholar.org/paper/4549649112553aabccfac8b918c7e98cdbdd0f09) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/4549649112553aabccfac8b918c7e98cdbdd0f09) | Automated discovery entry for Pashto resource tracking. |
+| Fairness Evaluation and Inference Level Mitigation in LLMs | [datacite](https://figshare.mq.edu.au/articles/thesis/Fairness_Evaluation_and_Inference_Level_Mitigation_in_LLMs/31093552/1) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://figshare.mq.edu.au/articles/thesis/Fairness_Evaluation_and_Inference_Level_Mitigation_in_LLMs/31093552/1) | Automated discovery entry for Pashto resource tracking. |
+| Fragments of life in ‘death world’: an analysis of Pashto poetry as a non-violent resistance to necropolitics | [other](https://www.semanticscholar.org/paper/9726f372b07f677fad23e2ee27a7f50f985e8ed8) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/9726f372b07f677fad23e2ee27a7f50f985e8ed8) | Automated discovery entry for Pashto resource tracking. |
+| Framing Political Bias in Multilingual LLMs Across Pakistani Languages | [datacite](https://arxiv.org/abs/2506.00068) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://arxiv.org/abs/2506.00068) | Automated discovery entry for Pashto resource tracking. |
 | From Scarcity to Scale: A Release-Level Analysis of the Pashto Common Voice Dataset | [arxiv](http://arxiv.org/abs/2602.14062v1) | [Matched by Pashto marker in paper title from arXiv query results. (`pashto`)](http://arxiv.org/abs/2602.14062v1) | ASR data quality and release trend reference |
+| From Scarcity to Scale: A Release-Level Analysis of the Pashto Common Voice Dataset | [datacite](https://arxiv.org/abs/2602.14062) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://arxiv.org/abs/2602.14062) | Automated discovery entry for Pashto resource tracking. |
+| Gemination in Pashto | [crossref](https://doi.org/10.24312/ucp-jll.02.02.405) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.24312/ucp-jll.02.02.405) | Automated discovery entry for Pashto resource tracking. |
+| Gender Classification From Pashto Handwritten Text Images | [other](https://www.semanticscholar.org/paper/2d70fffa9224d71f67ad3c1943b8a71b18164eeb) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/2d70fffa9224d71f67ad3c1943b8a71b18164eeb) | Automated discovery entry for Pashto resource tracking. |
+| Introduction to Pashto Word’s Characteristics | [other](https://www.semanticscholar.org/paper/6eb3febbb368a7eaccc6290bcd77683ed3d624aa) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/6eb3febbb368a7eaccc6290bcd77683ed3d624aa) | Automated discovery entry for Pashto resource tracking. |
+| Is the Pushto a Semitic Language | [zenodo](https://zenodo.org/records/16001185) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/16001185) | Automated discovery entry for Pashto resource tracking. |
+| Isolated Handwritten Pashto Character Recognition Using a <i>K</i>‐NN Classification Tool based on Zoning and HOG Feature Extraction Techniques | [openalex](https://doi.org/10.1155/2021/5558373) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1155/2021/5558373) | Automated discovery entry for Pashto resource tracking. |
 | KNN and ANN-based Recognition of Handwritten Pashto Letters using Zoning Features | [arxiv](http://arxiv.org/abs/1904.03391v2) | [Matched by arXiv query: all:pashto. (`pashto`)](http://arxiv.org/abs/1904.03391v2) | Pashto research reference for methods and benchmarking |
+| KNN and ANN-based Recognition of Handwritten Pashto Letters using Zoning Features | [openalex](https://doi.org/10.14569/ijacsa.2018.091069) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.14569/ijacsa.2018.091069) | Automated discovery entry for Pashto resource tracking. |
 | KPTI: Katib's Pashto Text Imagebase and Deep Learning Benchmark | [openalex](https://doi.org/10.1109/icfhr.2016.0090) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icfhr.2016.0090) | Pashto OCR dataset and benchmarking reference |
+| Language Barrier and its Effect on Learning at the Public Primary School Level in Lahore | [zenodo](https://zenodo.org/records/17728944) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/17728944) | Automated discovery entry for Pashto resource tracking. |
+| Le verbe pashto | [crossref](https://doi.org/10.29091/9783954907083) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.29091/9783954907083) | Automated discovery entry for Pashto resource tracking. |
+| Morphology of Pashto Adverbs: Word and Paradigm Approach | [zenodo](https://zenodo.org/records/16211508) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/16211508) | Automated discovery entry for Pashto resource tracking. |
+| Negation in Pashto | [zenodo](https://zenodo.org/records/18233956) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/18233956) | Automated discovery entry for Pashto resource tracking. |
+| Negotiating Pakhto: Proverbs, Islam and the Construction of Identity among Pashtuns | [other](https://www.semanticscholar.org/paper/8a503f164e0c1f5be13866dad00539c7e5b1cabc) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/8a503f164e0c1f5be13866dad00539c7e5b1cabc) | Automated discovery entry for Pashto resource tracking. |
+| Only 2 of 141 Global Languages Employ a Labial for "Tongue" in 1st position Challenging Saussure's Arbitrariness With Near Universal Embodied Iconicity for Tongue Vs Mouth in "inverse" Control | [datacite](https://zenodo.org/doi/10.5281/zenodo.17807676) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/doi/10.5281/zenodo.17807676) | Automated discovery entry for Pashto resource tracking. |
 | Out-of-Vocabulary Pashto Spell Checker using Morphological Operations | [other](https://www.semanticscholar.org/paper/802aae68a6a7fdfb29d51be03fb2b09e29311fa7) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/802aae68a6a7fdfb29d51be03fb2b09e29311fa7) | Pashto research reference for methods and benchmarking |
+| Pashto | [crossref](https://doi.org/10.4324/9780203301524-34) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.4324/9780203301524-34) | Automated discovery entry for Pashto resource tracking. |
+| Pashto (Endo-)clitics in a Parallel Architecture | [openalex](http://nbn-resolving.de/urn:nbn:de:bsz:352-0-278290) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](http://nbn-resolving.de/urn:nbn:de:bsz:352-0-278290) | Automated discovery entry for Pashto resource tracking. |
+| Pashto : Pashto-English, English-Pashto dictionary & phrasebook | [other](https://www.semanticscholar.org/paper/8ff77d35396d17225d97772e577e472a2ab1c47a) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/8ff77d35396d17225d97772e577e472a2ab1c47a) | Automated discovery entry for Pashto resource tracking. |
+| Pashto free relatives and triply-filled Comp: Evidence for a headed analysis | [openalex](https://doi.org/10.1016/s0024-3841(96)00032-0) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1016/s0024-3841(96)00032-0) | Automated discovery entry for Pashto resource tracking. |
+| Pashto Handwritten Books | [crossref](https://doi.org/10.1163/9789004737358_003) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.1163/9789004737358_003) | Automated discovery entry for Pashto resource tracking. |
 | Pashto isolated digits recognition using deep convolutional neural network | [openalex](https://doi.org/10.1016/j.heliyon.2020.e03372) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1016/j.heliyon.2020.e03372) | Pashto speech recognition research reference |
+| Pashto Language | [crossref](https://doi.org/10.32388/pxbtfv) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.32388/pxbtfv) | Automated discovery entry for Pashto resource tracking. |
 | Pashto offensive language detection: a benchmark dataset and monolingual Pashto BERT | [openalex](https://doi.org/10.7717/peerj-cs.1617) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.7717/peerj-cs.1617) | Pashto NLP toxicity detection benchmark and model reference |
+| PASHTO POETRY AND MILITANCY IN KHYBER PAKHTUNKHWA AFTER 9/11: THEMATIC ANALYSIS OF PASHTO POETRY IN RESISTING MILITANCY | [other](https://www.semanticscholar.org/paper/e81d4e7ac6cd7519643bf5d5c0bdfd9be554a8f2) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/e81d4e7ac6cd7519643bf5d5c0bdfd9be554a8f2) | Automated discovery entry for Pashto resource tracking. |
+| Pashto preverbs V | [other](https://www.semanticscholar.org/paper/1f59f22ae99379106b417186f3053c00b5fe391f) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/1f59f22ae99379106b417186f3053c00b5fe391f) | Automated discovery entry for Pashto resource tracking. |
+| Pashto preverbs, III. Compound verbs with preverb | [other](https://www.semanticscholar.org/paper/53eeae3a973d6bb72839e9304be13a0362c92242) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/53eeae3a973d6bb72839e9304be13a0362c92242) | Automated discovery entry for Pashto resource tracking. |
 | Pashto Shallow Parsing: A Deep Learning Approach | [other](https://www.semanticscholar.org/paper/be36455bb4eca60accb3e6866f345132f0dac1e5) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/be36455bb4eca60accb3e6866f345132f0dac1e5) | Pashto research reference for methods and benchmarking |
+| Pashto Tappa | [crossref](https://doi.org/10.4324/9781003604877-9) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.4324/9781003604877-9) | Automated discovery entry for Pashto resource tracking. |
+| Pashto Verse | [crossref](https://doi.org/10.1017/s0041977x00072700) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.1017/s0041977x00072700) | Automated discovery entry for Pashto resource tracking. |
+| Persian loanwords and calques in Pashto | [other](https://www.semanticscholar.org/paper/ed232f1c2abd6e6f8a49f04de8ac76bf922521ea) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/ed232f1c2abd6e6f8a49f04de8ac76bf922521ea) | Automated discovery entry for Pashto resource tracking. |
+| Persian, Urdu, and Pashto: A comparative orthographic analysis | [openalex](https://doi.org/10.1093/wsr/wsq005) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1093/wsr/wsq005) | Automated discovery entry for Pashto resource tracking. |
 | PHTI: Pashto Handwritten Text Imagebase for Deep Learning Applications | [openalex](https://doi.org/10.1109/access.2022.3216881) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/access.2022.3216881) | Pashto OCR dataset and benchmark reference |
 | Pioneer dataset and recognition of Handwritten Pashto characters using Convolution Neural Networks | [openalex](https://doi.org/10.1177/0020294020964826) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1177/0020294020964826) | Pashto handwritten character recognition reference |
+| Portrayal of Death in the Selected Poems of Abdul Ghani Khan and Emily Dickinson: A Comparative Thematic Analysis | [zenodo](https://zenodo.org/records/15046502) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/15046502) | Automated discovery entry for Pashto resource tracking. |
 | POS tagging of low-resource Pashto language: annotated corpus and BERT-based model | [other](https://www.semanticscholar.org/paper/1b2d5c896fec735483e8c8fb0a75e13125e08769) | [Matched by Semantic Scholar query: pashto. (`pashto`)](https://www.semanticscholar.org/paper/1b2d5c896fec735483e8c8fb0a75e13125e08769) | Pashto research reference for methods and benchmarking |
 | PsOCR: Benchmarking Large Multimodal Models for Optical Character Recognition in Low-resource Pashto Language | [other](https://www.semanticscholar.org/paper/d2743c0dcdbc65f5b46fcec2f0ba7cb379c4134f) | [Paper title explicitly references low-resource Pashto language OCR benchmarking. (`Pashto`, `OCR`)](https://www.semanticscholar.org/paper/d2743c0dcdbc65f5b46fcec2f0ba7cb379c4134f) | Pashto OCR research baseline and evaluation reference |
+| PsOCR: Benchmarking Large Multimodal Models for Optical Character Recognition in Low-resource Pashto Language | [datacite](https://arxiv.org/abs/2505.10055) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://arxiv.org/abs/2505.10055) | Automated discovery entry for Pashto resource tracking. |
+| PsOCR: Benchmarking Large Multimodal Models for Optical Character Recognition in Low-resource Pashto Language | [arxiv](http://arxiv.org/abs/2505.10055v2) | [Matched by Pashto marker in paper title from arXiv query results. (`pashto`)](http://arxiv.org/abs/2505.10055v2) | Automated discovery entry for Pashto resource tracking. |
 | Recognition of Pashto Handwritten Characters Based on Deep Learning | [openalex](https://doi.org/10.3390/s20205884) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.3390/s20205884) | Pashto OCR model reference for handwritten character recognition |
 | Recognizable units in Pashto language for OCR | [openalex](https://doi.org/10.1109/icdar.2015.7333963) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icdar.2015.7333963) | Pashto OCR preprocessing and unit-design reference |
 | Scale and rotation invariant OCR for Pashto cursive script using MDLSTM network | [openalex](https://doi.org/10.1109/icdar.2015.7333931) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icdar.2015.7333931) | Pashto OCR model architecture reference |
+| Scale and rotation invariant recognition of cursive Pashto script using SIFT features | [openalex](https://doi.org/10.1109/icet.2010.5638470) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icet.2010.5638470) | Automated discovery entry for Pashto resource tracking. |
+| Separating phonology from syntax: a reanalysis of Pashto cliticization | [openalex](https://doi.org/10.1017/s0022226700006952) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1017/s0022226700006952) | Automated discovery entry for Pashto resource tracking. |
 | Shape analysis of Pashto script and creation of image database for OCR | [openalex](https://doi.org/10.1109/icet.2009.5353160) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1109/icet.2009.5353160) | Pashto OCR dataset design and feature reference |
+| Some Remarks on the Origin of the Afghán People and Dialect and on the Connexion of the Pushto Language with the Zend and Pehlavi and the Hebrew | [zenodo](https://zenodo.org/records/16191315) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/16191315) | Automated discovery entry for Pashto resource tracking. |
 | Speech translation for low-resource languages: the case of Pashto | [openalex](https://doi.org/10.21437/interspeech.2005-723) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.21437/interspeech.2005-723) | Pashto speech translation and low-resource MT reference |
+| Summaries in Pashto | [crossref](https://doi.org/10.1097/01.wtf.0000437933.40809.39) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.1097/01.wtf.0000437933.40809.39) | Automated discovery entry for Pashto resource tracking. |
+| SWITCHING SELVES ONLINE:PASHTO-ENGLISH BILINGUALISM,IDENTITY, AND EXPRESSION IN PAKISTAN’S DIGITAL DISCOURSE | [other](https://www.semanticscholar.org/paper/7a330c5fb416a1105866a895748b4336f8ef8100) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/7a330c5fb416a1105866a895748b4336f8ef8100) | Automated discovery entry for Pashto resource tracking. |
+| Syntax and morphology of Baniswola Pashto: investigating universal and dialectal variations | [other](https://www.semanticscholar.org/paper/9f725b3b282cf05f9089002d474010c6021001f9) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/9f725b3b282cf05f9089002d474010c6021001f9) | Automated discovery entry for Pashto resource tracking. |
+| The BBN Byblos Pashto OCR system | [openalex](https://doi.org/10.1145/1031442.1031447) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1145/1031442.1031447) | Automated discovery entry for Pashto resource tracking. |
+| The development and evaluation of an automatic clitic generator for Pashto language | [other](https://www.semanticscholar.org/paper/3d95449d67799fcac83f855984cb0c29cc500d7b) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/3d95449d67799fcac83f855984cb0c29cc500d7b) | Automated discovery entry for Pashto resource tracking. |
+| The grammar of clitics : evidence from Pashto and other languages | [openalex](http://hdl.handle.net/11858/00-001M-0000-0012-7447-0) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](http://hdl.handle.net/11858/00-001M-0000-0012-7447-0) | Automated discovery entry for Pashto resource tracking. |
+| The Influence of the Arabic Language on the Pashto Language: The Abdur-Rahman Baba as a Model (A Case Study Analysis) | [zenodo](https://zenodo.org/records/18174368) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/18174368) | Automated discovery entry for Pashto resource tracking. |
+| The Pashto language and identity‐formation in Pakistan | [openalex](https://doi.org/10.1080/09584939508719759) | [Matched by explicit Pashto marker in title from OpenAlex works search. (`pashto`)](https://doi.org/10.1080/09584939508719759) | Automated discovery entry for Pashto resource tracking. |
+| The Regional Criminal Law Framework for the Combat of Organized Crime (Pashto) | [datacite](https://opendata.uni-halle.de//handle/1981185920/123975) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://opendata.uni-halle.de//handle/1981185920/123975) | Automated discovery entry for Pashto resource tracking. |
+| The Role of Early Literary Biographies (Tazkiri) in the Ancient History of Pashto Literature | [other](https://www.semanticscholar.org/paper/4938170077d3430c2e3f9fadc161ed7b79242917) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/4938170077d3430c2e3f9fadc161ed7b79242917) | Automated discovery entry for Pashto resource tracking. |
+| The Roshani Movement literary services and the contribution of this Movement in the development of Pashto Literature | [other](https://www.semanticscholar.org/paper/88a3cd1ec497844c5997ae1795f8e72bbb314112) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/88a3cd1ec497844c5997ae1795f8e72bbb314112) | Automated discovery entry for Pashto resource tracking. |
+| The Social Structure and Organization of A Pakhto Speaking Community in Afghanistan. | [other](https://www.semanticscholar.org/paper/306e9a04b8835de6e906303b5e27d43a6994cb1d) | [Matched by explicit Pashto marker in paper title from Semantic Scholar search. (`pashto`)](https://www.semanticscholar.org/paper/306e9a04b8835de6e906303b5e27d43a6994cb1d) | Automated discovery entry for Pashto resource tracking. |
+| Topicalization in Pashto | [crossref](https://doi.org/10.31703/gssr.2020(v-i).17) | [Matched by explicit Pashto marker in title from Crossref search. (`pashto`)](https://doi.org/10.31703/gssr.2020(v-i).17) | Automated discovery entry for Pashto resource tracking. |
 | Tuning Traditional Language Processing Approaches for Pashto Text Classification | [arxiv](http://arxiv.org/abs/2305.03737v1) | [Matched by Pashto marker in paper title from arXiv query results. (`pashto`)](http://arxiv.org/abs/2305.03737v1) | Pashto text classification method reference |
+| بلوچستان میں " فقہ اسلامی "   کے فروغ و ارتقا٫ کا تحقیقی جائزہ | [zenodo](https://zenodo.org/records/18049233) | [Zenodo metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/records/18049233) | Automated discovery entry for Pashto resource tracking. |
 
 ## Maintenance
 - Source of truth: [../catalog/resources.json](../catalog/resources.json)

resources/projects/README.md

@@ -4,20 +4,51 @@
 
 | Resource | Link | Pashto Evidence | Primary Use |
 |---|---|---|---|
+| adnankarim/ihanif-whisper-base-pashto | [huggingface](https://huggingface.co/spaces/adnankarim/ihanif-whisper-base-pashto) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/adnankarim/ihanif-whisper-base-pashto) | Automated discovery entry for Pashto resource tracking. |
+| adnankarim/ihanif-whisper-medium-pashto-3e-7 | [huggingface](https://huggingface.co/spaces/adnankarim/ihanif-whisper-medium-pashto-3e-7) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/adnankarim/ihanif-whisper-medium-pashto-3e-7) | Automated discovery entry for Pashto resource tracking. |
 | afaaaak/urdu_pashto_translator | [huggingface](https://huggingface.co/spaces/afaaaak/urdu_pashto_translator) | [Space metadata title is Urdu Pashto Translator and the slug includes pashto. (`Pashto`, `translator`)](https://huggingface.co/spaces/afaaaak/urdu_pashto_translator) | Translation demo and bilingual usability testing |
 | afaqalinagra/PASHTO-ASR-MODEL | [huggingface](https://huggingface.co/spaces/afaqalinagra/PASHTO-ASR-MODEL) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/afaqalinagra/PASHTO-ASR-MODEL) | Interactive Pashto demo and quick qualitative validation |
 | Aizazayyubi/pashto_asr | [huggingface](https://huggingface.co/spaces/Aizazayyubi/pashto_asr) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/Aizazayyubi/pashto_asr) | Interactive Pashto ASR demo for qualitative evaluation |
+| amirajorloo/jira-auto-direction-chrome-extension | [github](https://github.com/amirajorloo/jira-auto-direction-chrome-extension) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/amirajorloo/jira-auto-direction-chrome-extension) | Automated discovery entry for Pashto resource tracking. |
 | DrSaqlainHassan/PashtoTokenixer | [huggingface](https://huggingface.co/spaces/DrSaqlainHassan/PashtoTokenixer) | [Space card title states Pashto Parts of Speech Identifier and the slug contains Pashto. (`Pashto`, `parts-of-speech`)](https://huggingface.co/spaces/DrSaqlainHassan/PashtoTokenixer) | Pashto NLP demo for token and part-of-speech analysis |
+| Early Pregnancy Loss [Pashto] | [datacite](https://zenodo.org/doi/10.5281/zenodo.18325729) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/doi/10.5281/zenodo.18325729) | Automated discovery entry for Pashto resource tracking. |
+| Fazlullahmamond/hadith-collection-pashto | [github](https://github.com/Fazlullahmamond/hadith-collection-pashto) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/Fazlullahmamond/hadith-collection-pashto) | Automated discovery entry for Pashto resource tracking. |
 | Fazlullahmamond/Pashto-Typing | [github](https://github.com/Fazlullahmamond/Pashto-Typing) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/Fazlullahmamond/Pashto-Typing) | Interactive Pashto demo and quick qualitative validation |
+| Female Birth Control Part I [Pashto] | [datacite](https://zenodo.org/doi/10.5281/zenodo.18325040) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/doi/10.5281/zenodo.18325040) | Automated discovery entry for Pashto resource tracking. |
+| Female Birth Control Part II [Pashto] | [datacite](https://zenodo.org/doi/10.5281/zenodo.18325401) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/doi/10.5281/zenodo.18325401) | Automated discovery entry for Pashto resource tracking. |
+| Haroon-blip/khan-pukhtoon | [github](https://github.com/Haroon-blip/khan-pukhtoon) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/Haroon-blip/khan-pukhtoon) | Automated discovery entry for Pashto resource tracking. |
+| Haseeb-007/Pashto-sekho | [huggingface](https://huggingface.co/spaces/Haseeb-007/Pashto-sekho) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/Haseeb-007/Pashto-sekho) | Automated discovery entry for Pashto resource tracking. |
+| haseebjanhamraz/PashtoFonts | [github](https://github.com/haseebjanhamraz/PashtoFonts) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/haseebjanhamraz/PashtoFonts) | Automated discovery entry for Pashto resource tracking. |
+| Hassaankabir/Pashto_Malgaray | [huggingface](https://huggingface.co/spaces/Hassaankabir/Pashto_Malgaray) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/Hassaankabir/Pashto_Malgaray) | Automated discovery entry for Pashto resource tracking. |
 | ihanif/wav2vec-pashto-asr | [huggingface](https://huggingface.co/spaces/ihanif/wav2vec-pashto-asr) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/ihanif/wav2vec-pashto-asr) | Interactive Pashto demo and quick qualitative validation |
 | ihanif/wav2vec2-bert-pashto-asr | [huggingface](https://huggingface.co/spaces/ihanif/wav2vec2-bert-pashto-asr) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/ihanif/wav2vec2-bert-pashto-asr) | Interactive Pashto demo and quick qualitative validation |
+| ihanif/whisper-medium-pashto | [huggingface](https://huggingface.co/spaces/ihanif/whisper-medium-pashto) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/ihanif/whisper-medium-pashto) | Automated discovery entry for Pashto resource tracking. |
+| IhyaCommunity/Khushkhat-Extension | [github](https://github.com/IhyaCommunity/Khushkhat-Extension) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/IhyaCommunity/Khushkhat-Extension) | Automated discovery entry for Pashto resource tracking. |
 | ilyas02828/Pashto_Sign_Language | [huggingface](https://huggingface.co/spaces/ilyas02828/Pashto_Sign_Language) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/ilyas02828/Pashto_Sign_Language) | Interactive Pashto demo and quick qualitative validation |
+| Introduction to Postpartum Care for Refugee women [Pashto] | [datacite](https://zenodo.org/doi/10.5281/zenodo.18324878) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/doi/10.5281/zenodo.18324878) | Automated discovery entry for Pashto resource tracking. |
+| lecramyajiv/fonts-arabic-extra | [github](https://github.com/lecramyajiv/fonts-arabic-extra) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/lecramyajiv/fonts-arabic-extra) | Automated discovery entry for Pashto resource tracking. |
+| lecramyajiv/ttf-x2 | [github](https://github.com/lecramyajiv/ttf-x2) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/lecramyajiv/ttf-x2) | Automated discovery entry for Pashto resource tracking. |
 | mahmudaq/PashtoASRNMT1 | [huggingface](https://huggingface.co/spaces/mahmudaq/PashtoASRNMT1) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/mahmudaq/PashtoASRNMT1) | Interactive Pashto demo and quick qualitative validation |
+| mastermoo/pashto-quran | [github](https://github.com/mastermoo/pashto-quran) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/mastermoo/pashto-quran) | Automated discovery entry for Pashto resource tracking. |
+| MuhammadUllah7/PAKHTOONN | [github](https://github.com/MuhammadUllah7/PAKHTOONN) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/MuhammadUllah7/PAKHTOONN) | Automated discovery entry for Pashto resource tracking. |
+| nabeelest/pakhtoodle | [github](https://github.com/nabeelest/pakhtoodle) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/nabeelest/pakhtoodle) | Automated discovery entry for Pashto resource tracking. |
+| NanoNulla/lorem | [github](https://github.com/NanoNulla/lorem) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/NanoNulla/lorem) | Automated discovery entry for Pashto resource tracking. |
 | nasirkhansayyad/pashto-whisper-demo | [huggingface](https://huggingface.co/spaces/nasirkhansayyad/pashto-whisper-demo) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/nasirkhansayyad/pashto-whisper-demo) | Interactive Pashto demo and quick qualitative validation |
+| Negation in Pashto | [datacite](https://zenodo.org/doi/10.5281/zenodo.18233956) | [DataCite metadata includes Pashto markers in title or description. (`pashto`)](https://zenodo.org/doi/10.5281/zenodo.18233956) | Automated discovery entry for Pashto resource tracking. |
+| nisarmasid/NisAr-PakhtoOn | [github](https://github.com/nisarmasid/NisAr-PakhtoOn) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/nisarmasid/NisAr-PakhtoOn) | Automated discovery entry for Pashto resource tracking. |
+| omid/Persian-Log2Vis | [github](https://github.com/omid/Persian-Log2Vis) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/omid/Persian-Log2Vis) | Automated discovery entry for Pashto resource tracking. |
+| Pakhtoon9900/Pakhtoon- | [github](https://github.com/Pakhtoon9900/Pakhtoon-) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/Pakhtoon9900/Pakhtoon-) | Automated discovery entry for Pashto resource tracking. |
 | Pashto ASR Space | [huggingface](https://huggingface.co/spaces/ihanif/pashto-asr) | [Space ID includes pashto-asr and is returned by Hugging Face Pashto space search. (`pashto`, `asr`)](https://huggingface.co/api/spaces/ihanif/pashto-asr) | Live Pashto speech-to-text demo project |
 | Pashto ASR V3 Space | [huggingface](https://huggingface.co/spaces/ihanif/pashto-asr-v3) | [Space card title is Pashto ASR V3 and short description states Pashto ASR. (`Pashto`, `ASR`)](https://huggingface.co/api/spaces/ihanif/pashto-asr-v3) | Project demo for Pashto ASR user testing |
 | Pashto to English Dictionary Space | [huggingface](https://huggingface.co/spaces/EngrAamirBangash/Pashto2English-Dictionary) | [Space metadata title states Pashto to English Dictionary. (`Pashto`)](https://huggingface.co/api/spaces/EngrAamirBangash/Pashto2English-Dictionary) | Interactive bilingual lookup project |
 | Pashto Translator Space | [huggingface](https://huggingface.co/spaces/Umar4321/Pashto-Translator) | [Space title is Pashto Translator and description states Pashto to English and Urdu translation. (`Pashto`)](https://huggingface.co/api/spaces/Umar4321/Pashto-Translator) | Interactive translation project demo |
+| Pukhtoon203/PUKHTOON | [github](https://github.com/Pukhtoon203/PUKHTOON) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/Pukhtoon203/PUKHTOON) | Automated discovery entry for Pashto resource tracking. |
+| Pukhtoonmafia009/Pukhtoonmafia009 | [github](https://github.com/Pukhtoonmafia009/Pukhtoonmafia009) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/Pukhtoonmafia009/Pukhtoonmafia009) | Automated discovery entry for Pashto resource tracking. |
+| pukhtoonyar406/pukhtoonyar406 | [github](https://github.com/pukhtoonyar406/pukhtoonyar406) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/pukhtoonyar406/pukhtoonyar406) | Automated discovery entry for Pashto resource tracking. |
+| ShahZamanPatan/Pashto-Baran | [github](https://github.com/ShahZamanPatan/Pashto-Baran) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/ShahZamanPatan/Pashto-Baran) | Automated discovery entry for Pashto resource tracking. |
+| ShawAnonymouse/Pakhtoon | [github](https://github.com/ShawAnonymouse/Pakhtoon) | [Repository metadata (name/description/topics) includes Pashto markers. (`pashto`)](https://github.com/ShawAnonymouse/Pakhtoon) | Automated discovery entry for Pashto resource tracking. |
+| tasal9/pashto-base-bloom-space | [huggingface](https://huggingface.co/spaces/tasal9/pashto-base-bloom-space) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/tasal9/pashto-base-bloom-space) | Automated discovery entry for Pashto resource tracking. |
+| tasal9/ZamAI-mt5-Pashto-Demo | [huggingface](https://huggingface.co/spaces/tasal9/ZamAI-mt5-Pashto-Demo) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/tasal9/ZamAI-mt5-Pashto-Demo) | Automated discovery entry for Pashto resource tracking. |
 | tasal9/ZamAI-Phi3-Mini-Pashto-Demo | [huggingface](https://huggingface.co/spaces/tasal9/ZamAI-Phi3-Mini-Pashto-Demo) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/tasal9/ZamAI-Phi3-Mini-Pashto-Demo) | Interactive Pashto demo and quick qualitative validation |
 | Umar4321/Pashto-To-English-Urdu | [huggingface](https://huggingface.co/spaces/Umar4321/Pashto-To-English-Urdu) | [Matched by Pashto keyword in Hugging Face Spaces search. (`pashto`)](https://huggingface.co/spaces/Umar4321/Pashto-To-English-Urdu) | Interactive Pashto demo and quick qualitative validation |
 | ZamAI-Mistral-7B-Pashto Space | [huggingface](https://huggingface.co/spaces/tasal9/ZamAI-Mistral-7B-Pashto-space) | [Space title and ID explicitly include Pashto and model card metadata exposes project details. (`Pashto`)](https://huggingface.co/api/spaces/tasal9/ZamAI-Mistral-7B-Pashto-space) | Interactive Pashto LLM project demo |

scripts/README.md

@@ -9,6 +9,7 @@ Automation scripts for quality checks, resource catalog validation, and search i
 - `generate_resource_views.py`: generate `resources/*/README.md`, `resources/README.md`, and `docs/search/resources.json` from the catalog.
 - `sync_resources.py`: collect new candidate Pashto resources from Kaggle, Hugging Face (datasets/models/spaces), GitHub, GitLab, OpenAlex, Crossref, Zenodo, Dataverse, DataCite, arXiv, and Semantic Scholar into `resources/catalog/pending_candidates.json`.
 - `promote_candidates.py`: auto-promote valid non-duplicate entries from `pending_candidates.json` into `resources/catalog/resources.json`.
+- `review_existing_resources.py`: review current catalog resources, remove stale/removed entries only with strong reasons, and log removals in `resources/catalog/removal_log.json`.
 - `run_resource_cycle.py`: run the full repeatable resource cycle with one command.
 
 ## Usage
@@ -33,11 +34,26 @@ Sync candidate resources for maintainer review:
 python scripts/sync_resources.py --limit 20
 ```
 
+Review existing resources and remove stale entries before discovery:
+```bash
+python scripts/review_existing_resources.py
+```
+
+Run stricter relevance cleanup mode:
+```bash
+python scripts/review_existing_resources.py --enforce-pashto-relevance
+```
+
 Auto-promote valid candidates into verified catalog:
 ```bash
 python scripts/promote_candidates.py
 ```
 
+Auto-promote while skipping online URL availability checks:
+```bash
+python scripts/promote_candidates.py --skip-url-check
+```
+
 Run full repeatable cycle:
 ```bash
 python scripts/run_resource_cycle.py --limit 25

scripts/promote_candidates.py

@@ -18,8 +18,23 @@ try:
 except ModuleNotFoundError:
     from validate_resource_catalog import validate_resource
 
+try:
+    from scripts.review_existing_resources import probe_resource_url
+except ModuleNotFoundError:
+    from review_existing_resources import probe_resource_url
+
 
 PLACEHOLDER_PRIMARY_USE = "Needs maintainer review before promotion to verified catalog."
+NOT_FOUND_PATTERNS = (
+    "repository not found",
+    "model not found",
+    "dataset not found",
+    "space not found",
+    "page not found",
+    "not found",
+    "this repository does not exist",
+    "we couldn't find",
+)
 
 
 def _canonical_url(value: str) -> str:
@@ -43,11 +58,24 @@ def _prepare_candidate(candidate: dict[str, Any]) -> dict[str, Any]:
     return promoted
 
 
+def _candidate_url_unavailable(url: str, timeout: float) -> bool:
+    probe = probe_resource_url(url, timeout)
+    if probe.hard_missing:
+        return True
+    if probe.content_sample:
+        lowered = probe.content_sample.casefold()
+        if any(pattern in lowered for pattern in NOT_FOUND_PATTERNS):
+            return True
+    return False
+
+
 def promote_candidates(
     catalog: dict[str, Any],
     pending_payload: dict[str, Any],
     *,
     max_promotions: int | None = None,
+    verify_urls: bool = False,
+    url_timeout: float = 10.0,
 ) -> tuple[list[dict[str, Any]], dict[str, int]]:
     resources = catalog.get("resources")
     if not isinstance(resources, list):
@@ -69,7 +97,7 @@ def promote_candidates(
     }
 
     promoted: list[dict[str, Any]] = []
-    stats = {"total": len(candidates), "promoted": 0, "duplicate": 0, "invalid": 0}
+    stats = {"total": len(candidates), "promoted": 0, "duplicate": 0, "invalid": 0, "unavailable": 0}
 
     for candidate in candidates:
         if max_promotions is not None and len(promoted) >= max_promotions:
@@ -90,6 +118,10 @@ def promote_candidates(
             stats["duplicate"] += 1
             continue
 
+        if verify_urls and _candidate_url_unavailable(url, url_timeout):
+            stats["unavailable"] += 1
+            continue
+
         errors = validate_resource(resource, len(resources) + len(promoted))
         if errors:
             stats["invalid"] += 1
@@ -112,6 +144,8 @@ def main() -> int:
     parser.add_argument("--catalog", default="resources/catalog/resources.json")
     parser.add_argument("--candidates", default="resources/catalog/pending_candidates.json")
     parser.add_argument("--max-promotions", type=int, default=None)
+    parser.add_argument("--skip-url-check", action="store_true")
+    parser.add_argument("--url-timeout", type=float, default=10.0)
     args = parser.parse_args()
 
     catalog_path = Path(args.catalog)
@@ -135,18 +169,22 @@ def main() -> int:
         catalog,
         pending_payload,
         max_promotions=args.max_promotions,
+        verify_urls=not args.skip_url_check,
+        url_timeout=args.url_timeout,
     )
     if not promoted:
         print(
             "Promotion complete: no new verified resources "
-            f"(duplicates={stats['duplicate']}, invalid={stats['invalid']})"
+            f"(duplicates={stats['duplicate']}, invalid={stats['invalid']}, unavailable={stats['unavailable']})"
         )
         return 0
 
     catalog_path.write_text(json.dumps(catalog, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
     print(
         "Promotion complete: "
-        f"promoted={stats['promoted']} duplicate={stats['duplicate']} invalid={stats['invalid']}"
+        "promoted="
+        f"{stats['promoted']} duplicate={stats['duplicate']} invalid={stats['invalid']} "
+        f"unavailable={stats['unavailable']}"
     )
     return 0
 

scripts/review_existing_resources.py

@@ -0,0 +1,423 @@
+"""Review existing catalog entries and remove only with strong evidence.
+
+This script enforces a conservative pre-sync audit:
+- Keep resources that are reachable and Pashto-relevant.
+- Remove only when there is a strong reason (for example hard 404/410, duplicate ID/URL,
+  or no Pashto signal in metadata and live page content).
+- Persist removal reasons in a log for maintainer review.
+
+Usage:
+    python scripts/review_existing_resources.py
+    python scripts/review_existing_resources.py --timeout 15
+    python scripts/review_existing_resources.py --dry-run
+"""
+
+from __future__ import annotations
+
+import argparse
+import concurrent.futures as futures
+import json
+import re
+import socket
+from dataclasses import dataclass
+from datetime import date, datetime, timezone
+from pathlib import Path
+from typing import Any
+from urllib.error import HTTPError, URLError
+from urllib.request import Request, urlopen
+
+try:
+    from scripts.validate_resource_catalog import validate_resource
+except ModuleNotFoundError:
+    from validate_resource_catalog import validate_resource
+
+
+USER_AGENT = "pashto-resource-review/1.0"
+MAX_BODY_BYTES = 120_000
+HARD_REMOVE_HTTP_CODES = {404, 410, 451}
+NOT_FOUND_PATTERNS = (
+    "repository not found",
+    "model not found",
+    "dataset not found",
+    "space not found",
+    "page not found",
+    "not found",
+    "this repository does not exist",
+    "we couldn't find",
+)
+AUTOMATED_PRIMARY_USE = "Automated discovery entry for Pashto resource tracking."
+PASHTO_WORD_RE = re.compile(r"(?<![A-Za-z0-9])(pashto|pukhto|pushto|pakhto)(?![A-Za-z0-9])", re.IGNORECASE)
+PASHTO_CODE_RE = re.compile(r"\b(ps(_af)?|pus|pbt[_-]?arab)\b", re.IGNORECASE)
+PASHTO_SCRIPT_MARKERS = ("پښتو", "پشتو")
+
+
+@dataclass
+class UrlProbe:
+    status_code: int | None = None
+    final_url: str | None = None
+    content_sample: str = ""
+    hard_missing: bool = False
+    uncertain_error: str | None = None
+
+
+def _contains_pashto_marker(value: str) -> bool:
+    text = (value or "").strip()
+    if not text:
+        return False
+    if PASHTO_WORD_RE.search(text):
+        return True
+    if PASHTO_CODE_RE.search(text):
+        return True
+    return any(marker in text for marker in PASHTO_SCRIPT_MARKERS)
+
+
+def _resource_metadata_has_pashto_signal(resource: dict[str, Any]) -> bool:
+    values: list[str] = []
+    for key in ("title", "url", "summary", "primary_use"):
+        item = resource.get(key)
+        if isinstance(item, str):
+            values.append(item)
+
+    tags = resource.get("tags")
+    if isinstance(tags, list):
+        values.extend(tag for tag in tags if isinstance(tag, str))
+
+    evidence = resource.get("pashto_evidence")
+    if isinstance(evidence, dict):
+        for key in ("evidence_text", "evidence_url"):
+            item = evidence.get(key)
+            if isinstance(item, str):
+                values.append(item)
+        markers = evidence.get("markers")
+        if isinstance(markers, list):
+            values.extend(marker for marker in markers if isinstance(marker, str))
+
+    return any(_contains_pashto_marker(value) for value in values)
+
+
+def _resource_has_direct_pashto_signal(resource: dict[str, Any]) -> bool:
+    values: list[str] = []
+    for key in ("title", "url"):
+        item = resource.get(key)
+        if isinstance(item, str):
+            values.append(item)
+
+    evidence = resource.get("pashto_evidence")
+    if isinstance(evidence, dict):
+        evidence_url = evidence.get("evidence_url")
+        if isinstance(evidence_url, str):
+            values.append(evidence_url)
+        markers = evidence.get("markers")
+        if isinstance(markers, list):
+            values.extend(marker for marker in markers if isinstance(marker, str))
+
+    tags = resource.get("tags")
+    if isinstance(tags, list):
+        values.extend(tag for tag in tags if isinstance(tag, str))
+
+    return any(_contains_pashto_marker(value) for value in values)
+
+
+def _is_automated_candidate_like(resource: dict[str, Any]) -> bool:
+    rid = resource.get("id")
+    primary_use = resource.get("primary_use")
+    return (isinstance(rid, str) and rid.startswith("candidate-")) or (
+        isinstance(primary_use, str) and primary_use.strip() == AUTOMATED_PRIMARY_USE
+    )
+
+
+def _canonical_url(value: str) -> str:
+    return value.rstrip("/")
+
+
+def _request_url(url: str, method: str, timeout: float) -> UrlProbe:
+    request = Request(url, method=method, headers={"User-Agent": USER_AGENT})
+    try:
+        with urlopen(request, timeout=timeout) as response:
+            status = getattr(response, "status", 200)
+            final_url = response.geturl()
+            sample = ""
+            if method == "GET":
+                payload = response.read(MAX_BODY_BYTES)
+                sample = payload.decode("utf-8", errors="replace")
+            return UrlProbe(status_code=status, final_url=final_url, content_sample=sample)
+    except HTTPError as exc:
+        if method == "GET":
+            try:
+                payload = exc.read(MAX_BODY_BYTES)
+                sample = payload.decode("utf-8", errors="replace")
+            except Exception:  # noqa: BLE001
+                sample = ""
+        else:
+            sample = ""
+        return UrlProbe(
+            status_code=exc.code,
+            final_url=exc.geturl(),
+            content_sample=sample,
+            hard_missing=exc.code in HARD_REMOVE_HTTP_CODES,
+        )
+    except (URLError, TimeoutError, socket.timeout) as exc:
+        return UrlProbe(uncertain_error=str(exc))
+
+
+def probe_resource_url(url: str, timeout: float) -> UrlProbe:
+    head = _request_url(url, "HEAD", timeout)
+    if head.uncertain_error:
+        return head
+    if head.status_code in HARD_REMOVE_HTTP_CODES:
+        head.hard_missing = True
+        return head
+    if head.status_code in {403, 405, 429} or head.status_code is None:
+        get_result = _request_url(url, "GET", timeout)
+        if get_result.status_code in HARD_REMOVE_HTTP_CODES:
+            get_result.hard_missing = True
+        return get_result
+    if head.status_code and 200 <= head.status_code < 400:
+        get_result = _request_url(url, "GET", timeout)
+        if get_result.uncertain_error:
+            return head
+        return get_result
+    return head
+
+
+def review_resources(
+    catalog: dict[str, Any],
+    *,
+    timeout: float = 12.0,
+    enforce_pashto_relevance: bool = False,
+    max_workers: int = 12,
+    probe_fn: Any = probe_resource_url,
+) -> tuple[dict[str, Any], dict[str, Any]]:
+    resources = catalog.get("resources")
+    if not isinstance(resources, list):
+        raise ValueError("catalog.resources must be a list")
+
+    kept: list[dict[str, Any]] = []
+    removals: list[dict[str, Any]] = []
+    warnings: list[str] = []
+    seen_ids: dict[str, str] = {}
+    seen_urls: dict[tuple[str, str], str] = {}
+
+    probe_results: dict[str, UrlProbe] = {}
+    candidate_urls = sorted(
+        {
+            resource.get("url", "").strip()
+            for resource in resources
+            if isinstance(resource, dict) and isinstance(resource.get("url"), str) and resource.get("url", "").strip()
+        }
+    )
+    if candidate_urls:
+        worker_count = max(1, min(max_workers, len(candidate_urls)))
+        with futures.ThreadPoolExecutor(max_workers=worker_count) as executor:
+            future_map = {executor.submit(probe_fn, url, timeout): url for url in candidate_urls}
+            for future in futures.as_completed(future_map):
+                url = future_map[future]
+                try:
+                    probe_results[url] = future.result()
+                except Exception as exc:  # noqa: BLE001
+                    probe_results[url] = UrlProbe(uncertain_error=str(exc))
+
+    for index, resource in enumerate(resources):
+        if not isinstance(resource, dict):
+            removals.append(
+                {
+                    "id": f"resource-{index}",
+                    "title": "",
+                    "url": "",
+                    "reasons": ["Entry is not a JSON object."],
+                    "evidence": {},
+                }
+            )
+            continue
+
+        rid = resource.get("id", "")
+        title = resource.get("title", "")
+        url = resource.get("url", "")
+        category = resource.get("category", "")
+        reasons: list[str] = []
+
+        if not isinstance(rid, str) or not rid.strip():
+            reasons.append("Missing or invalid resource id.")
+        if not isinstance(url, str) or not url.strip():
+            reasons.append("Missing or invalid resource URL.")
+
+        if isinstance(rid, str) and rid:
+            previous = seen_ids.get(rid)
+            if previous:
+                reasons.append(f"Duplicate resource id; already used by '{previous}'.")
+
+        canonical_url = _canonical_url(url) if isinstance(url, str) else ""
+        normalized_category = str(category).strip().casefold() if isinstance(category, str) else ""
+        if canonical_url:
+            previous = seen_urls.get((normalized_category, canonical_url))
+            if previous:
+                reasons.append(
+                    "Duplicate canonical URL in same category; "
+                    f"already used by '{previous}'."
+                )
+
+        validation_errors = validate_resource(resource, index)
+        if any(".url must be a valid http/https URL" in error for error in validation_errors):
+            reasons.append("Resource URL failed schema validation.")
+
+        probe = UrlProbe()
+        if isinstance(url, str) and url.strip():
+            probe = probe_results.get(url, UrlProbe())
+            if probe.hard_missing:
+                status_code = probe.status_code if probe.status_code is not None else "unknown"
+                reasons.append(f"URL returned hard-missing HTTP status {status_code}.")
+            elif probe.uncertain_error:
+                warnings.append(f"{rid or f'resource-{index}'} URL probe inconclusive: {probe.uncertain_error}")
+
+        metadata_pashto = _resource_metadata_has_pashto_signal(resource)
+        direct_pashto = _resource_has_direct_pashto_signal(resource)
+        page_pashto = _contains_pashto_marker(probe.content_sample)
+        page_not_found = any(pattern in probe.content_sample.casefold() for pattern in NOT_FOUND_PATTERNS)
+
+        if page_not_found and not page_pashto:
+            reasons.append("Live page content indicates resource is unavailable.")
+
+        if enforce_pashto_relevance and not metadata_pashto and not page_pashto:
+            reasons.append("No Pashto signal found in metadata or live page content.")
+
+        if enforce_pashto_relevance and _is_automated_candidate_like(resource) and not direct_pashto and not page_pashto:
+            reasons.append("Automated candidate lacks direct Pashto signal and appears low-confidence.")
+
+        if reasons:
+            removals.append(
+                {
+                    "id": rid,
+                    "title": title,
+                    "url": url,
+                    "reasons": reasons,
+                    "evidence": {
+                        "status_code": probe.status_code,
+                        "final_url": probe.final_url,
+                        "metadata_pashto": metadata_pashto,
+                        "direct_pashto": direct_pashto,
+                        "page_pashto": page_pashto,
+                    },
+                }
+            )
+            continue
+
+        kept.append(resource)
+        if isinstance(rid, str) and rid:
+            seen_ids[rid] = title if isinstance(title, str) else rid
+        if canonical_url:
+            seen_urls[(normalized_category, canonical_url)] = (
+                title if isinstance(title, str) else canonical_url
+            )
+
+    updated_catalog = dict(catalog)
+    if len(kept) != len(resources):
+        updated_catalog["resources"] = kept
+        updated_catalog["updated_on"] = date.today().isoformat()
+
+    report = {
+        "checked": len(resources),
+        "kept": len(kept),
+        "removed": len(removals),
+        "removals": removals,
+        "warnings": warnings,
+    }
+    return updated_catalog, report
+
+
+def update_removal_log(log_path: Path, removals: list[dict[str, Any]]) -> None:
+    payload: dict[str, Any]
+    if log_path.exists():
+        try:
+            payload = json.loads(log_path.read_text(encoding="utf-8"))
+        except json.JSONDecodeError:
+            payload = {"updated_on": date.today().isoformat(), "entries": []}
+    else:
+        payload = {"updated_on": date.today().isoformat(), "entries": []}
+
+    entries = payload.get("entries")
+    if not isinstance(entries, list):
+        entries = []
+
+    removed_on = datetime.now(timezone.utc).isoformat()
+    for item in removals:
+        entries.append(
+            {
+                "removed_on": removed_on,
+                "id": item.get("id", ""),
+                "title": item.get("title", ""),
+                "url": item.get("url", ""),
+                "reasons": item.get("reasons", []),
+                "evidence": item.get("evidence", {}),
+            }
+        )
+
+    payload["updated_on"] = date.today().isoformat()
+    payload["entries"] = entries
+    log_path.parent.mkdir(parents=True, exist_ok=True)
+    log_path.write_text(json.dumps(payload, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--catalog", default="resources/catalog/resources.json")
+    parser.add_argument("--timeout", type=float, default=12.0)
+    parser.add_argument("--max-workers", type=int, default=12)
+    parser.add_argument("--removal-log", default="resources/catalog/removal_log.json")
+    parser.add_argument("--dry-run", action="store_true")
+    parser.add_argument(
+        "--enforce-pashto-relevance",
+        action="store_true",
+        help="Also remove entries that have no Pashto signal in metadata or live page content.",
+    )
+    args = parser.parse_args()
+
+    catalog_path = Path(args.catalog)
+    removal_log_path = Path(args.removal_log)
+    if not catalog_path.exists():
+        print(f"Missing catalog file: {catalog_path}")
+        return 1
+
+    try:
+        catalog = json.loads(catalog_path.read_text(encoding="utf-8"))
+    except json.JSONDecodeError as exc:
+        print(f"Invalid catalog JSON: {exc}")
+        return 1
+
+    updated_catalog, report = review_resources(
+        catalog,
+        timeout=args.timeout,
+        enforce_pashto_relevance=args.enforce_pashto_relevance,
+        max_workers=args.max_workers,
+        probe_fn=probe_resource_url,
+    )
+
+    print(
+        "Resource review complete: "
+        f"checked={report['checked']} kept={report['kept']} removed={report['removed']} "
+        f"warnings={len(report['warnings'])}"
+    )
+
+    if report["warnings"]:
+        for warning in report["warnings"]:
+            print(f"[warn] {warning}")
+
+    if report["removed"]:
+        for item in report["removals"]:
+            rid = item.get("id", "<unknown>")
+            reasons = item.get("reasons", [])
+            print(f"[remove] {rid}")
+            for reason in reasons:
+                print(f"  - {reason}")
+
+        if not args.dry_run:
+            catalog_path.write_text(json.dumps(updated_catalog, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+            update_removal_log(removal_log_path, report["removals"])
+    elif not args.dry_run and updated_catalog != catalog:
+        # Defensive branch for any non-removal edits.
+        catalog_path.write_text(json.dumps(updated_catalog, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

scripts/run_resource_cycle.py

@@ -9,6 +9,8 @@ Usage:
     python scripts/run_resource_cycle.py --skip-pytest
     python scripts/run_resource_cycle.py --discover-only
     python scripts/run_resource_cycle.py --max-promotions 10
+    python scripts/run_resource_cycle.py --skip-existing-review
+    python scripts/run_resource_cycle.py --skip-pashto-relevance-check
 """
 
 from __future__ import annotations
@@ -37,12 +39,38 @@ def main() -> int:
         default=None,
         help="Optional cap for auto-promotion count from pending candidates",
     )
+    parser.add_argument(
+        "--skip-existing-review",
+        action="store_true",
+        help="Skip review/removal of stale existing resources before syncing candidates.",
+    )
+    parser.add_argument(
+        "--resource-timeout",
+        type=float,
+        default=12.0,
+        help="Timeout in seconds for existing-resource URL probes.",
+    )
+    parser.add_argument(
+        "--skip-pashto-relevance-check",
+        action="store_true",
+        help="Disable Pashto relevance filtering in existing-resource review.",
+    )
     args = parser.parse_args()
 
     repo_root = Path(__file__).resolve().parents[1]
-    steps: list[list[str]] = [
-        ["python", "scripts/sync_resources.py", "--limit", str(args.limit)],
-    ]
+    steps: list[list[str]] = []
+    if not args.skip_existing_review:
+        review_step = [
+            "python",
+            "scripts/review_existing_resources.py",
+            "--timeout",
+            str(args.resource_timeout),
+        ]
+        if not args.skip_pashto_relevance_check:
+            review_step.append("--enforce-pashto-relevance")
+        steps.append(review_step)
+
+    steps.append(["python", "scripts/sync_resources.py", "--limit", str(args.limit)])
 
     if not args.discover_only:
         promote_step = ["python", "scripts/promote_candidates.py"]

tests/test_promote_candidates.py

@@ -1,5 +1,6 @@
 from datetime import date
 
+import scripts.promote_candidates as promote_module
 from scripts.promote_candidates import PLACEHOLDER_PRIMARY_USE, promote_candidates
 
 
@@ -132,3 +133,29 @@ def test_promote_candidates_respects_max_promotions() -> None:
     assert len(promoted) == 1
     assert stats["promoted"] == 1
     assert len(catalog["resources"]) == 2
+
+
+def test_promote_candidates_skips_unavailable_when_url_check_enabled(monkeypatch) -> None:
+    catalog = _catalog()
+    pending = {
+        "candidate_count": 1,
+        "candidates": [
+            _candidate(
+                rid="dataset-unavailable",
+                title="Pashto Unavailable Dataset",
+                url="https://example.org/pashto-unavailable",
+            )
+        ],
+    }
+
+    monkeypatch.setattr(
+        promote_module,
+        "_candidate_url_unavailable",
+        lambda *_args, **_kwargs: True,
+    )
+
+    promoted, stats = promote_candidates(catalog, pending, verify_urls=True)
+
+    assert promoted == []
+    assert stats["promoted"] == 0
+    assert stats["unavailable"] == 1

tests/test_review_existing_resources.py

@@ -0,0 +1,114 @@
+from scripts.review_existing_resources import UrlProbe, review_resources
+
+
+def _resource(*, rid: str, title: str, url: str) -> dict:
+    return {
+        "id": rid,
+        "title": title,
+        "url": url,
+        "category": "dataset",
+        "source": "other",
+        "status": "verified",
+        "summary": "Resource summary used for catalog review tests.",
+        "primary_use": "Testing",
+        "tasks": ["nlp"],
+        "pashto_evidence": {
+            "evidence_text": "Contains Pashto signal in metadata.",
+            "evidence_url": url,
+            "markers": ["Pashto"],
+        },
+        "tags": ["pashto", "dataset"],
+    }
+
+
+def test_review_resources_removes_hard_missing_urls() -> None:
+    catalog = {
+        "version": "1.0.0",
+        "updated_on": "2026-02-20",
+        "resources": [_resource(rid="dataset-a", title="Pashto A", url="https://example.org/a")],
+    }
+
+    def probe(_: str, __: float) -> UrlProbe:
+        return UrlProbe(status_code=404, hard_missing=True)
+
+    updated, report = review_resources(catalog, probe_fn=probe)
+
+    assert report["removed"] == 1
+    assert updated["resources"] == []
+    assert any("hard-missing HTTP status 404" in reason for reason in report["removals"][0]["reasons"])
+
+
+def test_review_resources_keeps_resource_when_probe_is_inconclusive() -> None:
+    catalog = {
+        "version": "1.0.0",
+        "updated_on": "2026-02-20",
+        "resources": [_resource(rid="dataset-a", title="Pashto A", url="https://example.org/a")],
+    }
+
+    def probe(_: str, __: float) -> UrlProbe:
+        return UrlProbe(uncertain_error="timed out")
+
+    updated, report = review_resources(catalog, probe_fn=probe)
+
+    assert report["removed"] == 0
+    assert len(updated["resources"]) == 1
+    assert len(report["warnings"]) == 1
+
+
+def test_review_resources_removes_duplicate_urls() -> None:
+    catalog = {
+        "version": "1.0.0",
+        "updated_on": "2026-02-20",
+        "resources": [
+            _resource(rid="dataset-a", title="Pashto A", url="https://example.org/shared"),
+            _resource(rid="dataset-b", title="Pashto B", url="https://example.org/shared"),
+        ],
+    }
+
+    def probe(_: str, __: float) -> UrlProbe:
+        return UrlProbe(status_code=200, content_sample="Pashto")
+
+    updated, report = review_resources(catalog, probe_fn=probe)
+
+    assert report["removed"] == 1
+    assert len(updated["resources"]) == 1
+    assert any("Duplicate canonical URL" in reason for reason in report["removals"][0]["reasons"])
+
+
+def test_review_resources_allows_same_url_across_different_categories() -> None:
+    dataset = _resource(rid="dataset-a", title="Pashto A", url="https://example.org/shared")
+    benchmark = _resource(rid="benchmark-a", title="Pashto A Benchmark", url="https://example.org/shared")
+    benchmark["category"] = "benchmark"
+    benchmark["tags"] = ["pashto", "benchmark"]
+    catalog = {
+        "version": "1.0.0",
+        "updated_on": "2026-02-20",
+        "resources": [dataset, benchmark],
+    }
+
+    def probe(_: str, __: float) -> UrlProbe:
+        return UrlProbe(status_code=200, content_sample="Pashto")
+
+    updated, report = review_resources(catalog, probe_fn=probe)
+
+    assert report["removed"] == 0
+    assert len(updated["resources"]) == 2
+
+
+def test_review_resources_enforces_pashto_relevance_only_when_enabled() -> None:
+    non_pashto = _resource(rid="dataset-x", title="General Dataset", url="https://example.org/general")
+    non_pashto["pashto_evidence"]["evidence_text"] = "Generic metadata note."
+    non_pashto["pashto_evidence"]["markers"] = ["generic"]
+    non_pashto["tags"] = ["dataset", "general"]
+    catalog = {"version": "1.0.0", "updated_on": "2026-02-20", "resources": [non_pashto]}
+
+    def probe(_: str, __: float) -> UrlProbe:
+        return UrlProbe(status_code=200, content_sample="General language resource")
+
+    updated_relaxed, report_relaxed = review_resources(catalog, probe_fn=probe, enforce_pashto_relevance=False)
+    updated_strict, report_strict = review_resources(catalog, probe_fn=probe, enforce_pashto_relevance=True)
+
+    assert report_relaxed["removed"] == 0
+    assert len(updated_relaxed["resources"]) == 1
+    assert report_strict["removed"] == 1
+    assert updated_strict["resources"] == []