# Resolve Production Incidents with an AI Workflow

Production incidents are difficult to diagnose across logs, code, dependencies, and systems. This skill structures investigation, repair, verification, rollout, and prevention across specialized AI agents.

## Install

```bash
npx skillstore add sickn33/incident-response-smart-fix
```

## Metadata

- Status: approved
- Slug: sickn33-incident-response-smart-fix
- Skillstore revision: r2
- Version status: missing
- Tree hash: cd0f0363a8431921cb2613dbbcd35507a1a0741e63e4b4b13d78d061c6bc6cd0
- Author: sickn33
- GitHub username: sickn33
- License: MIT
- Repository: https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/incident-response-smart-fix
- Ref: 88a8e9a07f4c54ab105c1c41b6267c287146b07b
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: external\_commands
- Quality score: 70
- Public page: https://skillstore.pages.dev/skills/sickn33-incident-response-smart-fix
- Manifest: https://skillstore.pages.dev/api/skills/sickn33-incident-response-smart-fix/manifest

## Capabilities

- Analyzes error signatures, stack traces, logs, traces, timelines, and user impact through structured prompts.
- Builds evidence-based root cause hypotheses using code paths, dependency checks, configuration comparisons, and git bisect.
- Routes work to language, database, performance, testing, and security specialists.
- Defines unit, integration, regression, security, cross-environment, and performance verification.
- Plans canary, blue-green, immediate, or feature-flag rollouts with rollback and monitoring criteria.
- Creates prevention tasks for documentation, linting, type safety, observability, and postmortems.

## Use Cases

- Triage a Production Failure: Organize error evidence, customer impact, recent changes, and immediate mitigation steps during an active incident.
- Coordinate a Cross-System Fix: Route database, application, performance, and testing work while preserving context between specialists.
- Strengthen Incident Prevention: Turn a resolved incident into tests, monitoring, static checks, documentation, and follow-up tasks.

## Prompt Templates

### Triage an Incident

```
Triage this incident: {summary}. Use {available_logs} and {known_impact}. Identify the error signature, affected users, timeline, and three immediate next steps.
```

### Investigate the Root Cause

```
Investigate the root cause of {incident}. Review {stack_trace}, {recent_changes}, and {dependency_context}. Produce evidence-backed hypotheses, reproduction steps, and a git bisect plan when appropriate.
```

### Design and Verify a Fix

```
Design a minimal fix for {root_cause} in {language_or_stack}. Include unit, integration, regression, security, and performance checks. Define rollback criteria and required monitoring.
```

### Coordinate a Complete Response

```
Coordinate a complete response for {incident} across {systems}. Assign specialist roles and specify context handoffs. Deliver a root cause, fix plan, verification matrix, rollout strategy, and prevention tasks.
```

## Limitations

- Does not connect to Sentry, DataDog, repositories, or deployment systems without separate tools and permissions.
- Referenced specialist agent names may not exist in every Claude, Codex, or Claude Code environment.
- Commands, thresholds, and templates are examples that require adaptation to the target system.
- Does not replace human approval for production changes or expert review of sensitive incidents.

## Best Practices

- Redact secrets and personal data before sharing logs, traces, links, or production samples.
- Confirm the incident scope, permissions, rollback plan, and success metrics before implementation.
- Validate fixes in controlled environments before a gradual production rollout.

## Anti Patterns

- Do not deploy generated fixes directly to production without authorized human review.
- Do not treat a plausible hypothesis as the root cause without reproduction or supporting evidence.
- Do not skip regression, security, or performance checks for high-impact incidents.

## Security Audit

- Audited at: 2026-07-24T01:09:02.559\+00:00
- Summary: All ten static findings are false positives. The matches are ordinary incident-response prose, Markdown resource links, and clear English rather than reconnaissance, shell execution, or obfuscation. Two medium-risk safeguards are missing: incident-data redaction rules and mandatory human approval before direct production deployment.

## Stats

- Views: 106
- Downloads: 9
- Favorites: 0
- Popularity score: 0