What problem does it solve?
When LOCI analyses fail, skills do not trigger, or results are missing or invalid, it is hard to gather the right environment and project state to diagnose the root cause; this Skill automates a thorough forensic snapshot and writes a sanitized, timestamped diagnostic report that can be shared for debugging.
Core Features & Use Cases
- Environment snapshot: Captures claude version, model ID, plugin version, OS info, project context, LOCI paths, setup markers, git state, and hooks configuration.
- 12-point diagnostics checklist: Runs checks for MCP visibility, session context, compiler and architecture detection, venv and asm-analyze health, build artifacts, hooks validity, and quota usage with clear PASS/FAIL details.
- Safe local forensics: Uses only local Read and shell operations (Read, Bash, Glob, Grep), avoids MCP tool use, sanitizes secrets and home paths, and writes a redacted markdown report for sharing or reloading into Claude Code.
- Use cases: Run after a failed analysis, when a skill did not auto-invoke, when MCP is unreachable, or whenever a developer needs a complete, reproducible failure snapshot.
Quick Start
Ask the plugin "bug report" followed by a one-sentence description of what failed.