agentleFS
tethertotetherto/qvac/.agents/skills/qv-pr-mine/SKILL.md

qv-pr-mine

Skill697 starsChanged 18 days agoApache-2.0View on GitHub

Show the current user's open PRs across every pod registered under .github/teams/, grouped by merge readiness, with copy-paste Slack ping messages routed to the owning pod's team. Use when the user asks about their own PRs, merge readiness, who to ping, or invokes /qv-pr-mine.

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run the script with --mode my.

    line 36
  2. Very specific:

    Slack-handle validation gate: If the script's stderr contains SLACK_VALIDATION_REQUIRED <N>, run the validation flow before presenting any …

    line 38

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

My PRs

Cross-pod skill. Surfaces the current user's open PRs in the configured repo and emits copy-paste Slack ping messages. The owning pod for each PR is auto-detected from the touched files against every .github/teams/<pod>.json's ownedPaths, so per-PR ping logic uses the right team automatically — no --pod argument needed.

When to use this skill

Use when:

  • User asks about their own PRs, merge readiness, or who to ping
  • User wants Slack messages to request reviews
  • User invokes /qv-pr-mine

Prerequisites

  • gh CLI installed and authenticated (gh auth status)
  • User must have access to the configured GitHub repository
  • At least one pod metadata file at .github/teams/<pod>.json
  • Per-user Slack handle map at ~/.config/qvac-pr-skills/slack.json (auto-bootstrapped on first run; see workflow step 2)

Usage

node .agents/skills/_lib/pr-skills/pr-status.mjs --mode my

--pod is intentionally omitted: --mode my discovers every pod and routes each PR to its owning pod automatically.

Workflow

  1. Run the script with --mode my. The script uses .agents/skills/_lib/pr-skills/pr-activity.mjs for shared PR state collection and ~/.config/qvac-pr-skills/config.json for repo/stale settings when available.
  2. Slack-handle validation gate. If the script's stderr contains SLACK_VALIDATION_REQUIRED <N>, run the validation flow before presenting any output to the user:

    a. Read ~/.config/qvac-pr-skills/slack.json. b. For each login in pendingReview, present the proposed handle to the user via AskQuestion so they can confirm or correct. Use one question per pending login. Show the GitHub login and the proposed handle as the option label, e.g. <github-login> -> <proposed-handle>. Provide options: "keep proposed", "edit (then prompt for new handle)". Do NOT inline any names from the proposed map into commentary, examples, or follow-up text. c. For any logins the user chose to edit, ask one follow-up question per login for the corrected handle as free text. d. Apply corrections to state.map, set state.pendingReview = [], save the file (atomic write). e. Re-run the script. The marker should no longer fire.

  3. Present the grouped output to the user (stderr contains progress info — ignore it).
  4. For PRs in "needs re-review" or "awaiting review", present the "Slack messages (copy-paste ready)" sections in copy-friendly fenced code blocks so the user can paste directly into Slack.
  5. Offer: "Want me to review any of these before requesting reviews? Provide the PR URL and I'll run /qv-pr-review."

Output groups

  1. Ready to merge — has both team member and team lead approval (per the PR's owning pod's team)
  2. Needs re-review — a reviewer's approval was dismissed (new commits); shows who to re-request
  3. Awaiting review — missing approvals; shows who to ping
  4. No pod matched — touched files don't match any .github/teams/<pod>.json ownedPaths; ping logic is skipped (rare; usually means the user's PR is outside the pod system)

Each non-empty group of PRs needing pings includes ready-to-copy Slack messages with @-mentions and PR links sourced from ~/.config/qvac-pr-skills/slack.json (falling back to @<github-login> when a handle is not yet mapped).

Per-PR pod resolution

For each of the user's PRs, the script:

  1. Lists the touched files via the GitHub API.
  2. Walks the discovered pods (one per .github/teams/*.json) and picks the first one whose ownedPaths contains any touched path as a prefix. The walk order is readdir order of .github/teams/.
  3. Uses that pod's leads, members, and Slack handles for the PR's "needs approval" / "ping" semantics.

If a PR touches files spanning multiple pods, the first match wins. This is fine for the typical case where a PR is owned by a single team. If two pods commonly share files, rename one of them in .github/teams/ so the desired pod sorts first.

Maintaining the Slack map

The script auto-fills new entries from gh api users/<login> when:

  • the file does not yet exist (first run on this machine), OR
  • a new login appears in any .github/teams/<pod>.json that is not yet in the map.

Newly seeded logins land in pendingReview and trigger the validation gate above on the next --mode my run. Edit ~/.config/qvac-pr-skills/slack.json directly at any time to update handles; the file is per-user and never committed.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.