# SkillPatch skill: gh-review-requests This skill fetches unread GitHub notifications for open, unmerged pull requests where a review is requested from a specified team or opened by a team member. It uses the GitHub CLI and a Python script (via `uv`) to filter PRs by team slug, then presents results in a formatted markdown table. A manual fallback using raw `gh api` calls is also provided. You (the agent) were given this URL and asked to use this skill. This is a **public** skill — no SkillPatch account, API key, or prior setup is required. Two ways to use it: **1. Use it right now** — the skill's complete file tree (SKILL.md + all reference files) is inlined below; read `SKILL.md` first, then follow it, consulting the other files as it directs. **2. Install the exact package onto disk** (recommended if you can run a shell — this reproduces the skill byte-for-byte, including any binary assets that can't be inlined): ```bash mkdir -p .claude/skills/gh-review-requests curl -sSL https://skillpatch.dev/install_skill/gh-review-requests | tar -xz -C .claude/skills/ ``` (`.claude/skills/` is Claude Code's convention; use whatever directory your agent loads skills from.) --- ## Skill files (2) - `SKILL.md` - `scripts/fetch_review_requests.py` ### `SKILL.md` ````markdown --- name: gh-review-requests description: Fetch unread GitHub notifications for open PRs where review is requested from a specified team or opened by a team member. Use when asked to "find PRs I need to review", "show my review requests", "what needs my review", "fetch GitHub review requests", or "check team review queue". allowed-tools: Bash --- # GitHub Review Requests Fetch unread `review_requested` notifications for open (unmerged) PRs, filtered by a GitHub team. **Requires**: GitHub CLI (`gh`) authenticated. **Requires**: The `uv` CLI for python package management, install guide at https://docs.astral.sh/uv/getting-started/installation/ ## Step 1: Identify the Team If the user has not specified a team, ask: > Which GitHub team should I filter by? (e.g. `streaming-platform`) Accept either a team slug (`streaming-platform`) or a display name ("Streaming Platform") — convert to lowercase-hyphenated slug before passing to the script. ## Step 2: Run the Script ```bash uv run scripts/fetch_review_requests.py --org getsentry --teams ``` To filter by multiple teams, pass a comma-separated list: ```bash uv run scripts/fetch_review_requests.py --org getsentry --teams ``` ### Script output ```json { "total": 3, "prs": [ { "notification_id": "12345", "title": "feat(kafka): add workflow to restart a broker", "url": "https://github.com/getsentry/ops/pull/19144", "repo": "getsentry/ops", "pr_number": 19144, "author": "bmckerry", "reasons": ["opened by: bmckerry"] } ] } ``` `reasons` will contain one or both of: - `"review requested from: "` — the team is a requested reviewer - `"opened by: "` — the PR author is a team member ## Step 3: Present Results Display results as a markdown table with full URLs: | # | Title | URL | Reason | |---|-------|-----|--------| | 1 | feat(kafka): add workflow to restart a broker | https://github.com/getsentry/ops/pull/19144 | opened by: evanh | If `total` is 0, say: "No unread review requests found for that team." ## Fallback If the script fails, run manually: ```bash gh api notifications --paginate ``` Then for each `review_requested` notification, check: - `gh api repos/{repo}/pulls/{number}` — skip if `state == "closed"` or `merged_at` is set - `gh api repos/{repo}/pulls/{number}/requested_reviewers` — check `teams[].name` - `gh api orgs/{org}/teams/{slug}/members` — check if author is a member ```` ### `scripts/fetch_review_requests.py` ``` # /// script # requires-python = ">=3.12" # dependencies = [] # /// """ Fetch unread GitHub review-requested notifications for open (unmerged) PRs, filtered by team membership and/or team review requests. Usage: uv run fetch_review_requests.py --org ORG --teams TEAM1,TEAM2 Arguments: --org GitHub organization slug (default: getsentry) --teams Comma-separated team slugs to filter by (e.g. streaming-platform) Output: JSON to stdout """ import argparse import json import subprocess import sys def gh(path: str, paginate: bool = False) -> list | dict: cmd = ["gh", "api", path] if paginate: cmd.append("--paginate") result = subprocess.run(cmd, capture_output=True, text=True) if result.returncode != 0 or not result.stdout: print(f"Error running gh {' '.join(cmd)}: {result.stderr}", file=sys.stderr) return [] if paginate else {} return json.loads(result.stdout) def main(): parser = argparse.ArgumentParser() parser.add_argument("--org", default="getsentry") parser.add_argument("--teams", required=True, help="Comma-separated team slugs") args = parser.parse_args() team_slugs = [t.strip() for t in args.teams.split(",")] # Resolve team members for all specified teams members: set[str] = set() team_display_names: dict[str, str] = {} for slug in team_slugs: data = gh(f"orgs/{args.org}/teams/{slug}/members", paginate=True) for m in data: members.add(m["login"]) # Get display name team_data = gh(f"orgs/{args.org}/teams/{slug}") team_display_names[slug] = team_data.get("name", slug) # Fetch unread notifications (GitHub API default: unread only) all_notifs = gh("notifications", paginate=True) review_notifs = [ n for n in all_notifs if n["reason"] == "review_requested" and n["unread"] ] prs = [] for n in review_notifs: url = n["subject"]["url"] repo_path = url.replace("https://api.github.com/repos/", "") repo = repo_path.rsplit("/pulls/", 1)[0] pr_num = repo_path.rsplit("/", 1)[-1] html_url = f"https://github.com/{repo}/pull/{pr_num}" pr_data = gh(f"repos/{repo}/pulls/{pr_num}") if not pr_data: continue # Skip merged or closed PRs if pr_data.get("merged_at") or pr_data.get("state") == "closed": continue author = pr_data["user"]["login"] reviewers_data = gh(f"repos/{repo}/pulls/{pr_num}/requested_reviewers") requested_team_names = [t["slug"] for t in reviewers_data.get("teams", [])] matching_teams = [ t for t in requested_team_names if any(slug.lower() == t.lower() for slug in team_slugs) ] by_team_member = author in members review_from_team = len(matching_teams) > 0 if not (by_team_member or review_from_team): continue reasons = [] if review_from_team: reasons.append(f"review requested from: {', '.join(matching_teams)}") if by_team_member: reasons.append(f"opened by: {author}") prs.append({ "notification_id": n["id"], "title": n["subject"]["title"], "url": html_url, "repo": repo, "pr_number": int(pr_num), "author": author, "reasons": reasons, }) print(json.dumps({"total": len(prs), "prs": prs}, indent=2)) if __name__ == "__main__": main() ```