CI failure triage
Turn red builds into actionable summaries
Watch CI, capture failing jobs, pull the relevant logs, and post a plain-language explanation of what broke and the most likely cause.
What it does
- ✓Detects failing CI runs as they happen
- ✓Pulls the relevant log excerpt for each failure
- ✓Explains the likely cause in plain language
- ✓Flags flaky vs real failures over time
The prompt
Drop this into a BrainClaw scheduled workflow. Adjust the schedule, sources, and thresholds to fit your setup.
You are my CI-triage agent. Trigger on CI failure webhooks: 1. Identify the failing job and pull the relevant log lines. 2. Summarize what broke in plain language and the most likely cause. 3. Note whether this test has failed flakily before. 4. Post to the team channel with a link to the run.
Run this recipe
Deploy an always-on BrainClaw agent and schedule it in minutes.
More recipes like CI failure triage
Release watch
Babysit a deploy so you don’t have to
Monitor a deploy or long build, retry on transient failures, and report the final outcome with logs when it finishes.
Daily GitHub digest
Wake up to issues triaged and PRs summarized
Pull new and updated issues, categorize by severity, flag anything assigned to you, and list open PRs with CI and review state — delivered as a morning summary.
PR review triage
Never let a pull request go stale
Track open pull requests, enforce review SLAs, nudge stale reviews, and summarize what each PR changes so reviewers start with context.
Inbox triage
Keep your inbox clear without lifting a finger
Read new email, label by priority, draft replies for routine threads, and surface anything that needs you — straight to your channel.