registry-bootstrap/.github/ISSUE_TEMPLATE/new-recipe.md

metadata

name: 💡 Propose new recipe
about: Suggest a new X-NN recipe for the TAF Agent
title: '[Proposal] X-NN — '
labels: recipe-proposed

Recipe number suggestion

X-__ (next free number — see existing recipes in taf_browser.py RECIPES dict)

What practical question does this answer

In one sentence: "Should/can/will I __ ?"

Formula chain

Which sections of the paper does it use?

• §17.X
• §19.X
• §26.X

Step-by-step formula chain (in pseudocode):

1. Compute X using §X.Y
2. Compute Y using §X.Z
3. If X > threshold → ...
4. Decide based on (X, Y) → verdict

Inputs needed

• input_1: __ (from where? config.json field? user-provided?)
• input_2: __

Output verdict types

What discrete verdicts can the recipe produce?

• YES / NO / DEGRADED
• Or specific labels like "USE SOFT DECAY" / "USE HARD CUTOFF"

Example use case

Walk through one realistic scenario where a technician would use this.

Existing literature

Are there papers / blog posts that motivate this recipe? Link them.