Watch Prs

Watch PRs

Start a /pr-status loop that runs until 18:00 local time, pacing itself adaptively: checking again soon (~3 min) while CI is running or a PR just changed, and backing off (10 → 30 min) once everything is settled. Unlike /watch-release, this loop is unattended — /pr-status only reads and renders (tables + suggested-action footer), it never prompts or blocks, so you can leave it running in a tab while you work.

Usage

/watch-prs            # adaptive cadence, stop at 18:00   (default)
/watch-prs 17         # adaptive, stop at 17:00
/watch-prs 10m        # FIXED 10-minute interval, stop at 18:00
/watch-prs 5m 17      # fixed 5m, stop at 17:00

Adaptive is the default. Pass an explicit \d+m interval only when you deliberately want a fixed glance rhythm (e.g. a tab you check on a known cadence).

Instructions

Parse optional arguments:

1. Interval — first arg matching \d+m. Absent → adaptive mode (the default). Present →

fixed-interval mode at that interval.

1. Stop hour — first arg matching \d{1,2} not already consumed as the interval. Default: 18.

If the current time is already past the stop hour, tell the user and don't start the loop.

Adaptive mode (no interval given)

Invoke the /loop skill in dynamic (self-paced) mode with /pr-status:

/loop /pr-status

Then pace each next wake from the recommendation /pr-status prints last in its tick output:

next-tick: {hot|warm|cold} (~{N}s) — {reason}

Use {N} as the delaySeconds for the next wake (the dynamic loop clamps to [60, 3600]). Don't second-guess the bucket — /pr-status already weighs CI / push / transition state (its step 6):

• hot (~180s) — CI in flight or a PR changed this tick; check soon to catch the result.
• warm (~600s) — PRs awaiting review, nothing time-critical.
• cold (1200 → 1800s) — settled; escalating back-off via the quietStreak counter.

Stop and don't reschedule once the wake would land past {stop_hour}:00. If a tick produces no next-tick: line (e.g. it errored before step 6), fall back to ~600s and continue.

Keep the tail terse. Each tick already ends with /pr-status's machine-readable next-tick: line, and the dynamic /loop prints its own one-line narrator for the next wake. Those two are the entire cadence budget — do NOT add a third prose line restating the bucket, CI state, or next-check time. The tables, deltas, and suggested-actions footer are the content; the cadence is one line.

Fixed mode (interval given)

Invoke the /loop skill with the literal interval — /pr-status's next-tick: line is ignored:

/loop {interval} /pr-status

Tell the loop to stop at {stop_hour}:00 local time.

Note

/pr-status is read-only and never prompts, so this loop runs unattended — leave it in a tab. The suggested-action footer points you at /ready-to-merge, /review-comments, etc. as PRs become actionable, but you run those yourself when you choose; the watcher never acts. For an attended release-gatekeeper that prompts to push/defer/cancel, use /watch-release instead.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: flurdy
• Source: flurdy/agent-skills
• License: MIT

Install and usage instructions live in the source repository linked above.