heartbeat-staleness-gate

Solid

Detect stale Squad heartbeat and require visible progress updates.

AI & Automation 58 stars 15 forks Updated 2 weeks ago MIT

Install

View on GitHub

Quality Score: 85/100

Stars 20%
59
Recency 20%
90
Frontmatter 20%
70
Documentation 15%
37
Issue Health 10%
50
License 10%
100
Description 5%
100

Skill Content

## Context Use when Ralph checks status, after major work batches, or before claiming long-running work is complete. ## Patterns - Read `.squad/heartbeat.md` frontmatter. - `updated_at`, `agent`, `current_task`, `last_action`, and `status` should reflect the latest meaningful state. - Rewrite frontmatter instead of appending duplicate frontmatter blocks. - If heartbeat is stale, route a small update to Scribe/Ralph before continuing with status claims. ## Examples - After a gate passes, set `status: committing`. - After work finishes, set `status: idle` and record `last_action`. ## Anti-Patterns - Do not leave stale "implementing" status after work has completed. - Do not bury current status only in agent output.

Details

Author
DanWahlin
Repository
DanWahlin/ai-agent-board
Created
6 months ago
Last Updated
2 weeks ago
Language
TypeScript
License
MIT

Integrates with

OpenAI · AI

Similar Skills

Semantically similar based on skill content — not just same category

AI & Automation Listed

today

Morning heartbeat — staleness checks, deadline surfacing, and memory curation for your workspace state. Run at the start of each day or after a 12+ hour gap.

2 Updated 1 months ago
conorbronsdon
AI & Automation Solid

heartbeat

Run agent heartbeat — periodic checks, memory consolidation, proactive work. Triggers on /agent:heartbeat, "heartbeat", "heartbeat check", "periodic check", "consolidar memoria".

62 Updated today
crisandrews
AI & Automation Listed

heartbeat

Give a scheduled job a dead-man's switch, so its silence is noticed — derive the grace from the job's own cadence, create the check through the healthchecks.io API, put the ping URL where the job's other credentials live, wire the exit codes onto the right endpoints, and prove it by watching the alarm fire. Also audits which scheduled jobs have no check at all. TRIGGER when: a scheduled job (backup, self-check, sync, report) has nothing that would notice it stopping; a heartbeat or dead-man's switch is being added, moved or diagnosed; a healthchecks.io check must be created or its grace chosen; someone asks how a job's failure would be noticed, or which jobs are unmonitored. DO NOT TRIGGER when: the task is provisioning the backup itself (that is `backup`, which hands over here at its last step); alerting for a web service's uptime rather than a scheduled job; or choosing where a secret lives (that is `doppler`).

1 Updated 2 days ago
AnotherSava