whats-on-fire

Ranks production failures, security exposure, and stalled work across every repo in a GitHub org.

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/Sassy-Dog/sassydog-skills --skill whats-on-fire-sassy-dog
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: whats-on-fire
Source: https://github.com/Sassy-Dog/sassydog-skills/tree/main/skills/whats-on-fire
Command: npx skills add https://github.com/Sassy-Dog/sassydog-skills --skill whats-on-fire-sassy-dog

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve? Engineering leads managing many repositories cannot see which product is actually broken today without manually checking Sentry, CI, Dependabot, and open PRs repo by repo. This Skill performs one read-only cross-product sweep of a GitHub org and returns a ranked, routed report of everything on fire. ## Core Features & Use Cases - Org-wide incident sweep: Pulls Sentry issues and cron monitors, org-wide open PRs and issues, failing workflows, Dependabot alerts, and code/secret scanning state, then ranks them P0/P1 across products. - Structural blind-spot detection: Flags repos with no error monitoring, no alerting, disabled dependency scanning, or Dependabot alerting without fix PRs. - Cron recovery cross-reference: Distinguishes a genuinely broken cron monitor from one whose fix was verified via workflow dispatch but awaits its next scheduled check-in. - Use Case: Ask "what's on fire across our products?" and receive a markdown report with production fires, security exposure, stuck shipping, backlog heat, and a cross-product top 5, each routed to its owning repo. ## Quick Start Ask the agent "what's on fire today across the org?" to run the full cross-product sweep and get a ranked report.

Frequently Asked Questions about whats-on-fire

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I check what is broken across all repos in a GitHub org?

Run this skill to sweep Sentry issues, cron monitors, failing workflows, Dependabot alerts, and open PRs across every active repo in the org. It ranks findings P0/P1 and routes each one to the owning repository in a single markdown report.

How does the skill rank Dependabot security alerts?

Dependabot exposure is ranked by remediation state, not alert count. A green fix PR sitting unmerged for 3+ days is P0, as are high/critical alerts aged 14+ days with no fix PR; fresh advisories with fixes already queued are reported as healthy.

Does the skill work without the GitHub CLI installed?

Yes, it falls back to the session's GitHub MCP tools using the cloud-fallback reference, rebuilding the roster and per-repo pulls via capability-resolved tools. The Dependabot, code-scanning, and secret-scanning surfaces have no MCP equivalent and are reported as named skips.

Can this skill file GitHub issues or change Sentry status?

No, the skill is strictly read-only by design. It never files issues, changes Sentry status, or mutates anything; it names the owning repo and the flow that would file, leaving that action to a human in the per-repo skills.

Why does a red Sentry cron monitor not always rank as P0?

An error-state monitor is downgraded when a green workflow_dispatch run completed after the last failing check-in, meaning the fix is verified and awaiting the next scheduled run. Monitors with an open auto-managed tracking issue are reported as working with a tracked finding instead.