Diagnose Claude Code session incidents and execute corrective actions.
npx skills add https://github.com/hgflima/loopy --skill heal-hgflima
Please help me install this Agent Skill. Skill: heal Source: https://github.com/hgflima/loopy/tree/main/.claude/skills/heal Command: npx skills add https://github.com/hgflima/loopy --skill heal-hgflima
The 'heal' Skill addresses post-incident analysis in Claude Code sessions, enabling quick diagnosis, corrective action, and structured changelog persistence to prevent recurrence.
Use the heal skill to analyze the incident reported in the current session and apply the proposed fix.
High-intent search queries and answers about installing and using this skill.
Post-mortem analysis diagnoses Claude Code session incidents by mining conversation history for root causes. The skill uses agents to formulate solutions, execute corrective actions on skills and rules, and persist an audit trail.
Yes, self-healing fixes for CLAUDE.md files are automated by analyzing the conversation history to formulate a corrective action. Agents execute the proposed fix and update the changelog to prevent future recurrence.
Logging root-cause incident analysis results is handled by creating a structured changelog. This persists the incident context, executed corrective actions, and audit information, providing a reference for future analysis and prevention.
No additional dependencies are required to run incident-analysis and corrective actions. The skill operates independently using its internal scripts, references, and assets to diagnose issues and execute fixes.
Limitations include relying entirely on available conversation history for root-cause identification. Corrective actions target skills, rules, and CLAUDE.md files, meaning unresolved external dependencies outside the session cannot be self-healed.
Yes, collaborative review is supported before executing session fixes. The skill's agents perform root cause analysis and solution formulation, followed by a collaborative review phase before any corrective action is applied to the environment.
imMamdouhaboammarAggregate GitHub, Jira, and Claude Code activity into structured standup updates.
AraflyOrganize incident facts into an evidence-based root cause analysis document.
RYN6666999Analyzes code dependencies to assess the impact and safety of proposed changes.
shekerkammaMonitor live production deploys for console errors, performance regressions, and page failures.
