RoyLee1224 commented on code in PR #69308: URL: https://github.com/apache/airflow/pull/69308#discussion_r3524846895
########## dev/skill-evals/README.md: ########## @@ -0,0 +1,129 @@ +<!-- + Licensed to the Apache Software Foundation (ASF) under one + or more contributor license agreements. See the NOTICE file + distributed with this work for additional information + regarding copyright ownership. The ASF licenses this file + to you under the Apache License, Version 2.0 (the + "License"); you may not use this file except in compliance + with the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, + software distributed under the License is distributed on an + "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY + KIND, either express or implied. See the License for the + specific language governing permissions and limitations + under the License. +--> + +<!-- START doctoc generated TOC please keep comment here to allow auto update --> +<!-- DON'T EDIT THIS SECTION, INSTEAD RE-RUN doctoc TO UPDATE --> +**Table of Contents** *generated with [DocToc](https://github.com/thlorenz/doctoc)* + +- [Skill-Eval Harness](#skill-eval-harness) + - [Prerequisites](#prerequisites) + - [One-time setup](#one-time-setup) + - [Usage](#usage) + - [Adding cases](#adding-cases) + - [How it works](#how-it-works) + - [Prek hook](#prek-hook) + +<!-- END doctoc generated TOC please keep comment here to allow auto update --> + +# Skill-Eval Harness + +Test whether changes to `AGENTS.md` break or improve agent decisions. +The harness compares the `main` branch version against your working +tree, running the same cases against both and reporting the diff. + +Each arm is a **git worktree** of the real repo — the agent sees +actual source files (`pyproject.toml`, directory structure, etc.). +The only difference between arms is which `AGENTS.md` is present. + +The agent reads guidance through the `CLAUDE.md → AGENTS.md` symlink, +so the harness verifies that symlink exists (in the working tree and +on the base branch) before running and aborts if it is broken — a +regular-file CLAUDE.md would make every arm read identical guidance. + +## Prerequisites + +- **Node.js >=22.22.0** — check with `node --version` +- **Authentication** (one of): + - Claude Code session (`claude /login`) — Pro/Max subscription + - `ANTHROPIC_API_KEY` environment variable — API credits + +## One-time setup + +Install the Claude Agent SDK (promptfoo needs it at runtime): + +```bash +mkdir -p ~/.promptfoo-sdk && cd ~/.promptfoo-sdk && npm init -y && npm install @anthropic-ai/claude-agent-sdk +``` + +## Usage + +```bash +# Check for regressions after editing AGENTS.md: +uv run dev/skill-evals/eval.py + +# Repeat each case to reduce nondeterminism: +uv run dev/skill-evals/eval.py --repeat 3 + +# Add baseline arm (no AGENTS.md) to measure raw model capability: +uv run dev/skill-evals/eval.py --full + +# Test a skill alongside AGENTS.md (not combinable with --full — the +# baseline arm has no skill, so the skill-used assertion would always fail): +SKILL_NAME=airflow-contribution uv run dev/skill-evals/eval.py + +# Use a cheaper model for fast iteration: +MODEL=claude-haiku-4-5-20251001 uv run dev/skill-evals/eval.py + +# Disable cache (force fresh LLM calls): +uv run dev/skill-evals/eval.py --no-cache + +# View results in browser (use the pinned version printed by eval.py): +npx [email protected] view Review Comment: Thanks for the compile-ui-assets-dev pointer ! `view-skill-eval` is now available prek run view-skill-eval --hook-stage manual --all-files -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
