| name | pr |
| description | toolkit for llm-friendly interaction with GitHub pull requests; using this skill the LLM can fetch PR data and analyze it |
| metadata | {"short-description":"LLM-friendly interactions with GitHub PRs"} |
Import pull request data as markdown
Use this skill to fetch PR metadata and file diffs into LLM-friendly Markdown files.
Prereqs
- Ensure
gh is installed and gh auth login has been completed.
- Confirm setup:
python3 .codex/skills/pr/scripts/get_pr_data.py ensure-setup returns zero exit code.
Usage
Fetch a PR by number:
python3 .codex/skills/pr/scripts/get_pr_data.py get-pr 354
Fetch from a specific repo:
python3 .codex/skills/pr/scripts/get_pr_data.py get-pr 354 --repo tonkeeper/ios_private
Write to a custom output directory:
python3 .codex/skills/pr/scripts/get_pr_data.py get-pr 354 --output /tmp/pr-354
Output
- Default directory:
.context/tasks/<PR_NUMBER>/ (relative to repo root)
- One markdown file per diff (
0.md, 1.md, ...) with keys as headings and the patch field in a diff code block.
sha.json with base commit metadata is written to the same directory.
.env is loaded from the repo root if present.