investigate-crash

Investigate Crashlytics crash spikes by analyzing commits and producing post-mortem documents.

Updated Apr 21, 2026
One-click install
npx skills add https://github.com/Zeyad-37/tech-agency --skill investigate-crash-zeyad-37
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: investigate-crash
Source: https://github.com/Zeyad-37/tech-agency/tree/main/.claude/skills/investigate-crash
Command: npx skills add https://github.com/Zeyad-37/tech-agency --skill investigate-crash-zeyad-37

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When a crash spike hits production, teams need a fast, structured way to find the culprit commit, decide between a targeted fix and a rollback, and document the incident so lessons become tracked work instead of forgotten notes. ## Core Features & Use Cases - Culprit Identification: Cross-references the crash timeline with recent git commits, diffs, and blame output to pinpoint the change that introduced the crash. - Fix and Rollback Recommendations: Proposes a targeted code fix with exact file and line, plus a revert command against the last known-good commit. - Post-Mortem Generation: Writes a severity-classified post-mortem to docs/artifacts/post-mortem/ and appends a row to the shared INDEX.md. - Prevention Task Tracking: Converts every prevention action point into a board task with owner, priority, and due date via the board adapter. - Use Case: Crashlytics alerts that crash-free sessions dropped after yesterday's release. Run the investigation to identify the offending KMP commit, get a revert command, and file a post-mortem with tracked follow-up tasks. ## Quick Start Ask the agent to investigate the current Crashlytics crash spike and produce a post-mortem with a recommended fix.

Frequently Asked Questions about investigate-crash

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

FAQPage Schema
How do I investigate a Crashlytics crash spike with git history?▼

Cross-reference the crash spike start time with recent commits using git log filtered by date, then inspect suspicious commits with git show and git blame on the crashing file. Prioritize commits touching code paths in the stack trace or shared KMP code.

What should a crash post-mortem document include?▼

A crash post-mortem includes the incident title and date, severity classification (P0/P1/P2 based on affected sessions), a timeline from detection to fix, root cause with commit references, impact estimate, resolution, prevention action points with owners, and lessons learned.

What if there is no Crashlytics context file available?▼

If .claude/crashlytics-context.md does not exist, the investigation asks you directly for the stack trace, affected platform, approximate spike start time, and crash count. The git-based culprit analysis then proceeds with that information.

When should I rollback instead of shipping a targeted crash fix?▼

Rollback with git revert is preferred when the culprit commit is clearly identified, the fix is non-trivial, and user impact is high severity. A targeted fix is better when the root cause is well understood and the change is small and verifiable.

How are post-mortem prevention actions tracked after the investigation?▼

Every prevention action point becomes a board task in the Backlog column with an owner, priority matching the post-mortem, and a due date based on severity (P0 within 48 hours, P1 within one week). Recurring incidents get an additional P0 escalation task.