fix: llm.chat IndexError (lazy Ollama warm + safe _resolve_backend fallback) + chat self-send returns direct

README.md

@@ -97,13 +97,13 @@ Get HearthNet for your platform:
 | **macOS Desktop** | `python build/quickstart.py macos` | .app | ~200MB | Native macOS app bundle |
 | **Python (Any OS)** | [Source](https://github.com/ckal/HearthNet) | Python | - | `python app.py` - full mesh node |
 | **Docker** | [Dockerfile](Dockerfile) | Container | 2GB | `docker run -p 7860:7860 hearthnet:latest` |
-| **Guides & Docs** | [BUILD_GUIDE.md](BUILD_GUIDE.md) | Markdown | - | How to build for each platform |
+| **Guides & Docs** | [BUILD_GUIDE.md](docs/guides/BUILD_GUIDE.md) | Markdown | - | How to build for each platform |
 
 **Recommended Paths:**
 - 🚀 **Fastest** (5 min): PWA Web App - instant, no install
 - 💻 **Desktop** (3 min): Download EXE/AppImage and run
 - 🐳 **Server**: Docker container deployment
-- 📚 See [BUILD_GUIDE.md](BUILD_GUIDE.md) for detailed instructions
+- 📚 See [BUILD_GUIDE.md](docs/guides/BUILD_GUIDE.md) for detailed instructions
 
 ---
 
@@ -142,7 +142,7 @@ python app.py
 # 4. Tap menu → "Install app" or "Add to Home screen"
 ```
 
-**📱 Full Android Setup Guide:** [ANDROID_DEPLOYMENT_GUIDE.md](ANDROID_DEPLOYMENT_GUIDE.md)
+**📱 Full Android Setup Guide:** [ANDROID_DEPLOYMENT_GUIDE.md](docs/guides/ANDROID_DEPLOYMENT_GUIDE.md)
 - ✅ PWA (instant, no build)
 - 🔧 Native APK (optional, advanced)
 
@@ -430,7 +430,7 @@ python -m pytest tests/ --ignore=tests/test_e2e_user_stories.py -q  # skip Playw
 
 | Resource | Link |
 |----------|------|
-| Architecture | [ARCHITECTURE.md](ARCHITECTURE.md) |
+| Architecture | [ARCHITECTURE.md](docs/ARCHITECTURE.md) |
 | System overview | [docs/00-OVERVIEW.md](docs/00-OVERVIEW.md) |
 | Capability contract | [docs/CAPABILITY_CONTRACT.md](docs/CAPABILITY_CONTRACT.md) |
 | Roadmap | [docs/roadmap.md](docs/roadmap.md) |
@@ -454,7 +454,7 @@ HearthNet has **comprehensive test coverage** across all modules:
 | **Transport** (X01) | 69 | ✅ Enhanced (12→55%+) | [X01 Enhanced](tests/test_x01_enhanced.py) |
 | **Phase 2/3 Specs** (M14-M32, X05-X09) | 216 | 🏗️ Scaffolded | [P2/P3 Tests](tests/test_m1[4-9]_spec.py), [X0[5-9]](tests/test_x0[5-9]_spec.py) |
 | **Reference Docs** | 80 | 🏗️ Scaffolded | [API Contract](tests/test_capability_contract.py), [Glossary](tests/test_glossary.py) |
-| **Total** | **932** | **100% Pass** | [Full Report](TEST_SUITE_REPORT.md) |
+| **Total** | **932** | **100% Pass** | [Full Report](docs/reports/TEST_SUITE_REPORT.md) |
 
 **Code Coverage: 44% (6,043 / 10,743 lines)**
 - Well-covered (>70%): Identity, Bus, Types, UI core
@@ -501,7 +501,7 @@ git commit -m "my changes"  # ← tests run first!
 
 ### Test Documentation
 
-- **Full Report:** [TEST_SUITE_REPORT.md](TEST_SUITE_REPORT.md) — 783 tests, all modules
+- **Full Report:** [TEST_SUITE_REPORT.md](docs/reports/TEST_SUITE_REPORT.md) — 783 tests, all modules
 - **Coverage Enhancement:** [COVERAGE_ENHANCEMENT_REPORT.md](COVERAGE_ENHANCEMENT_REPORT.md) — 149 new tests for M04/M05/X01/X03
 - **Test Structure:** Each test follows Happy / Error / Edge pattern
 - **No Mocks:** All implemented tests use real code paths
@@ -525,7 +525,7 @@ git commit -m "my changes"  # ← tests run first!
 **Build Artifacts Available:**
 - Windows EXE: [dist/HearthNet.exe](https://huggingface.co/spaces/build-small-hackathon/HearthNet/resolve/main/dist/HearthNet.exe) (212 MB)
 - Android APK: [build/android/.../app-debug.apk](https://huggingface.co/spaces/build-small-hackathon/HearthNet/resolve/main/build/android/HearthNetApp/platforms/android/app/build/outputs/apk/debug/app-debug.apk) (3.56 MB)
-- Build scripts: [BUILD_GUIDE.md](BUILD_GUIDE.md) for EXE, AppImage, .app, Docker
+- Build scripts: [BUILD_GUIDE.md](docs/guides/BUILD_GUIDE.md) for EXE, AppImage, .app, Docker
 
 ---
 

docs/ARCHITECTURE.md

@@ -437,5 +437,5 @@ python -m hearthnet.conformance.runner --output conformance-report/
 ---
 
 *This document is generated from the spec set in `docs/`. For per-module detail see:*
-- *Phase 1+2: `docs/00-OVERVIEW.md`, `docs/CAPABILITY_CONTRACT.md`, `docs/M01-*.md` …*
+- *Phase 1+2: `00-OVERVIEW.md`, `CAPABILITY_CONTRACT.md`, `modules/M01-*.md` …*
 - *Phase 3: `docs/p2_p3/IMPLEMENTATION_REFERENCE_p3.md`, `docs/p2_p3/M26-*.md` …*

docs/guides/ANDROID_DEPLOYMENT_GUIDE.md

@@ -177,10 +177,10 @@ FastAPI Backend (http://localhost:7860)
 
 ## Documentation Files
 
-- [PWA Implementation](docs/M08-ui.md) - Full web app details
-- [Cordova Build Guide](build/android/CORDOVA_BUILD_GUIDE.md) - Detailed native build
-- [APK Setup](build/android/SETUP_COMPLETE.md) - Project status
-- [Build Paths](build/android/BUILD_PATHS.md) - Decision guide
+- [PWA Implementation](../modules/M08-ui.md) - Full web app details
+- [Cordova Build Guide](../../build/android/CORDOVA_BUILD_GUIDE.md) - Detailed native build
+- [APK Setup](../../build/android/SETUP_COMPLETE.md) - Project status
+- [Build Paths](../../build/android/BUILD_PATHS.md) - Decision guide
 
 ---
 

docs/guides/BUILD_GUIDE.md

@@ -387,9 +387,9 @@ gpg --verify dist/HearthNet-*.AppImage.sig
 ## 📚 Related Documentation
 
 - [BUILD_STATUS.md](BUILD_STATUS.md) - Current artifact inventory
-- [ANDROID_DEPLOYMENT_GUIDE.md](../ANDROID_DEPLOYMENT_GUIDE.md) - APK + PWA setup
+- [ANDROID_DEPLOYMENT_GUIDE.md](ANDROID_DEPLOYMENT_GUIDE.md) - APK + PWA setup
 - [README.md](../README.md) - Main project readme
-- [Cordova Guide](CORDOVA_BUILD_GUIDE.md) - Android native build
+- [Cordova Guide](../../build/android/CORDOVA_BUILD_GUIDE.md) - Android native build
 
 ---
 

docs/guides/HOWTO.md

@@ -146,7 +146,7 @@ For nodes behind NAT or across the internet, configure a relay URL:
 relay_urls = ["https://your-relay.example.com"]
 ```
 
-See [docs/p2_p3/M15-relay-tier.md](p2_p3/M15-relay-tier.md).
+See [docs/p2_p3/M15-relay-tier.md](../p2_p3/M15-relay-tier.md).
 
 ### Checking connected peers
 
@@ -525,7 +525,7 @@ python -m hearthnet.cli call emergency.probe@1.0 '{}'
 
 ## 11. How Routing Works
 
-**Spec:** [docs/M03-bus.md](M03-bus.md) §3.5 / §5.4  
+**Spec:** [docs/modules/M03-bus.md](../modules/M03-bus.md) §3.5 / §5.4  
 **Implementation:** `hearthnet/bus/router.py`, `hearthnet/bus/__init__.py`
 
 ### The capability bus

BUILD_STATUS.md → docs/reports/BUILD_STATUS.md

@@ -82,9 +82,9 @@
   - Architecture diagrams
   
 - **Build Guides**:
-  - [Cordova Build Guide](build/android/CORDOVA_BUILD_GUIDE.md)
-  - [Build Paths Decision Guide](build/android/BUILD_PATHS.md)
-  - [Setup Complete Status](build/android/SETUP_COMPLETE.md)
+  - [Cordova Build Guide](../../build/android/CORDOVA_BUILD_GUIDE.md)
+  - [Build Paths Decision Guide](../../build/android/BUILD_PATHS.md)
+  - [Setup Complete Status](../../build/android/SETUP_COMPLETE.md)
 
 ---
 
@@ -253,9 +253,9 @@ adb install build/android/HearthNetApp/platforms/android/app/build/outputs/apk/d
 ## 📞 Support
 
 **Documentation**:
-- [Deployment Guide](ANDROID_DEPLOYMENT_GUIDE.md)
-- [Cordova Build Guide](build/android/CORDOVA_BUILD_GUIDE.md)
-- [Architecture Docs](docs/M08-ui.md)
+- [Deployment Guide](../guides/ANDROID_DEPLOYMENT_GUIDE.md)
+- [Cordova Build Guide](../../build/android/CORDOVA_BUILD_GUIDE.md)
+- [Architecture Docs](../modules/M08-ui.md)
 
 **GitHub**: [ckal/HearthNet](https://github.com/ckal/HearthNet)  
 **HF Space**: [build-small-hackathon/HearthNet](https://huggingface.co/spaces/build-small-hackathon/HearthNet)  

docs/upgrade_plan.md

@@ -128,8 +128,8 @@ and (when keys set) the Nemotron/Modal backends.
 
 ## Phase 7 — Documentation
 
-**Files:** `README.md`, `docs/M*.md` capability-status lines, `docs/GLOSSARY.md`,
-`docs/CAPABILITY_CONTRACT.md`.
+**Files:** `README.md`, `modules/M*.md` capability-status lines, `GLOSSARY.md`,
+`CAPABILITY_CONTRACT.md`.
 
 - Record the bge-small embedding model and that RAG is now real semantic retrieval.
 - **Model policy:** keep `SmolLM2-135M-Instruct` as the default LLM (tiny-titan track,

hearthnet/services/chat/service.py

@@ -52,7 +52,25 @@ class ChatService:
         recipient = payload.get("recipient") or payload.get("to", "")
 
         if recipient == self._node_id:
-            return {"error": "bad_request", "message": "Cannot send to self"}
+            # Self-message: store locally and return direct status
+            event_id = payload.get("event_id") or f"msg:{uuid.uuid4().hex}"
+            now = datetime.now(UTC).strftime("%Y-%m-%dT%H:%M:%SZ")
+            msg = {
+                "event_id": event_id,
+                "from": self._node_id,
+                "to": self._node_id,
+                "body": payload.get("body", ""),
+                "attachments": payload.get("attachments", []),
+                "sent_at": now,
+                "client_id": payload.get("client_id", event_id),
+            }
+            self.messages.append(msg)
+            if self._view is not None:
+                try:
+                    self._view._messages.append(msg)
+                except Exception:
+                    pass
+            return {"output": {"event_id": event_id, "lamport": 0, "delivered": "direct"}, "meta": {}}
 
         event_id = payload.get("event_id") or f"msg:{uuid.uuid4().hex}"
         client_id = payload.get("client_id", event_id)

hearthnet/services/llm/service.py

@@ -89,23 +89,37 @@ class LlmService:
         ]
 
     def _resolve_backend(self, model_name: str) -> tuple[LlmBackend, str]:
-        """Pick the backend that serves ``model_name``; fall back to primary."""
+        """Pick the backend that serves ``model_name``; fall back to first backend with models."""
         if model_name:
             for backend in self._backends:
                 for bm in backend.models:
                     if bm.name == model_name:
                         return backend, model_name
-        backend = self._backends[0]
-        return backend, backend.models[0].name
+        # Safe fallback: skip any backend whose model list is still empty (not yet warmed)
+        for backend in self._backends:
+            if backend.models:
+                return backend, backend.models[0].name
+        raise RuntimeError(
+            "No LLM backend has available models. "
+            "Run `ollama pull <model>`, start llama.cpp, or install transformers."
+        )
 
     async def _handle_chat(self, req: RouteRequest) -> dict:
+        # Lazy warm: populate model lists for backends that haven't been warmed yet
+        # (e.g. OllamaBackend.models is empty until warm() is awaited)
+        for _b in self._backends:
+            if not _b.models:
+                try:
+                    await _b.warm()
+                except Exception:
+                    pass
         inp = req.body.get("input", {})
         messages = inp.get("messages", [])
         params = req.body.get("params", {})
-        backend, model_name = self._resolve_backend(str(params.get("model") or ""))
         temperature = float(params.get("temperature", 0.7))
         max_tokens = int(params.get("max_tokens", 1024))
         try:
+            backend, model_name = self._resolve_backend(str(params.get("model") or ""))
             result = await backend.chat(
                 messages,
                 model=model_name,
@@ -126,11 +140,17 @@ class LlmService:
             return {"error": "internal_error", "message": str(exc)}
 
     async def _handle_complete(self, req: RouteRequest) -> dict:
+        for _b in self._backends:
+            if not _b.models:
+                try:
+                    await _b.warm()
+                except Exception:
+                    pass
         inp = req.body.get("input", {})
         prompt = inp.get("prompt", "")
         params = req.body.get("params", {})
-        backend, model_name = self._resolve_backend(str(params.get("model") or ""))
         try:
+            backend, model_name = self._resolve_backend(str(params.get("model") or ""))
             result = await backend.complete(prompt, model=model_name, stream=False)
             return {
                 "output": {"text": result.text},

tasks.md

@@ -5,12 +5,12 @@
 All Phase 1 (M01-M13, X01-X04), Phase 2 (M14-M25, X05-X07), and Phase 3 experimental
 (M26-M31) modules are implemented. **489 tests pass, 59 skipped (E2E), 0 fail**.
 
-See [ARCHITECTURE.md](ARCHITECTURE.md) for the full module map, data flows, and local-to-HF setup guide.
-See [docs/IMPROVEMENTS.md](docs/IMPROVEMENTS.md) for the full improvement backlog and prize targeting analysis.
+See [ARCHITECTURE.md](docs/ARCHITECTURE.md) for the full module map, data flows, and local-to-HF setup guide.
+See [docs/reports/IMPROVEMENTS.md](docs/reports/IMPROVEMENTS.md) for the full improvement backlog and prize targeting analysis.
 
 ---
 
-## Maximize Real Activation (June 12) — see [upgrade_plan.md](upgrade_plan.md)
+## Maximize Real Activation (June 12) — see [upgrade_plan.md](docs/upgrade_plan.md)
 
 Full 10-phase upgrade to make every feasible capability *genuinely real* (no mocks,
 fakes, or `# noqa`/`# nosec` bypasses). **Net result: +18 passing tests, −6 failing,
@@ -104,7 +104,7 @@ pre-existing `playwright` `ModuleNotFound` (optional browser-test dep).
   MiniCPM (MINICPM_URL), and Modal (MODAL_ENDPOINT) backends from env vars
 - README: added `nemotron`, `minicpm`, `modal` tags; expanded hackathon section
   with sponsor prize targeting table
-- `docs/IMPROVEMENTS.md`: comprehensive improvement backlog with GPT-4o rating,
+- `docs/reports/IMPROVEMENTS.md`: comprehensive improvement backlog with GPT-4o rating,
   29 improvement items, and priority matrix
 
 **README + submission (June 11):**