SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsbabysit-prs

babysit-prs

General & other

View repository

Updated 10-15 17:25

About this skill

Babysit PRs

Keep your open PRs moving without polling them by hand. Meant to run on a loop — /loop /babysit-prs — one pass per tick; --once runs a single pass.

The boundary, one line: act without asking only on the mechanical (a nit reply through an already-gated skill); every judgment (which P1/P2 to fix, the merge) is a STOP. This skill composes /pr-feedback and /pr-respond — it adds no new intelligence, only the scan and the waiting.

Each tick

  1. List your open PRs in the current repo:

    gh pr list --author "@me" --state open --json number,url,title

  2. Read each PR's standing state, then the delta. Two things, in this order — the delta alone is a trap: a backlog of unresolved threads that predates the loop never changes, so a delta-only tick reports — while comments pile up unanswered.

    a. Standing count (every tick, regardless of change). Per PR, count open review threads — isResolved == false and isOutdated == false — split by author type. One GraphQL call gives both the count and the authors:

gh api graphql -f query='query($owner:String!,$repo:String!,$pr:Int!){
  repository(owner:$owner,name:$repo){ pullRequest(number:$pr){
    reviewThreads(first:100){ nodes{ isResolved isOutdated
      comments(first:1){ nodes{ author{login __typename} } } } } } }
}' -F owner=OWNER -F repo=REPO -F pr=<n>

b. Mergeability + CI (every tick). gh pr view <n> --json mergeable,mergeStateStatus for the merge state, and gh pr checks <n> for the CI rollup — count conclusions (pass / fail / pending / skipping). Report the raw numbers, not a synthesized verdict.

c. Delta. Then diff against last tick — which of those threads, reviews, checks, or mergeable state are new since the previous pass.

Loop state lives in this conversation, not on disk. If a tick has no memory of the last one, re-triaging from scratch is safe: /pr-feedback is read-only, so a repeat pass costs tokens, not correctness. This is the waiting ceiling — accept it rather than building a state store.

Author type is a label, not a filter. Mark an author 🤖 when GraphQL author{ __typename } is Bot, OR the login ends in [bot], OR it's a known bot account (naboo-ai-reviews, cursor/Bugbot); everyone else is 👤. Never drop bot activity — bot findings are shown alongside human ones. This composes with /pr-feedback, which does no author typing of its own.

  1. Triage new activity through /pr-feedback (classifies P1 / P2 / Nit).

  2. Emit the tick report. One table, one row per PR — standing backlog and activity, not just state, so neither a fresh comment nor a pile of unanswered ones can hide behind —:

    Colonne Contenu
    PR #num + state glyph (✅ ready / ⚠️ problème / 🚧 draft/WIP)
    Mergeable mergeable + mergeStateStatus from step 2b, e.g. ✅ CLEAN, ⚠️ BEHIND (rebase), ⛔ BLOCKED, ❌ DIRTY (conflicts)
    CI check rollup from step 2b — surface failures first: ❌ 1 fail · 2 pending · 43 pass (skipping omitted unless it's all there is). ✅ all pass when green.
    À répondre standing count of open threads (step 2a), split e.g. 👤 2 · 🤖 6; 0 when the PR is clean. This is the number that persists across ticks — never — just because nothing changed.
    Nouveau delta only — new comments/reviews this tick, each 👤 <login> or 🤖 <login> + ≤5-word gist; — when nothing new since last tick.
    P1/P2/Nit count of open actionable items from the /pr-feedback triage (the table links to the triage, it doesn't replace it)

    Then act:

    • Nits and clear agreements → batch through /pr-respond (mechanical — it already gates on one confirmation and runs replies through humanizer).
    • Which P1 / P2 to actually fix → surface it and STOP. That choice is the user's.

  3. Notify on the events worth coming back for, via PushNotification (load it first: ToolSearch "select:PushNotification"):

    • an APPROVE → "merge-ready."
    • all checks green / PR ready for human review.
    • After a push answering feedback: notify "re-solicit the reviewer" — do not re-request review on GitHub and do not ping Slack. The user re-solicits.

Cadence

Launch with /loop /babysit-prs (~20 min is plenty for review latency). In dynamic mode, tighten the interval when a PR is close to green and loosen it otherwise via ScheduleWakeup.