What problem does it solve? When a chat agent behaves unexpectedly—running slowly, skipping tools, failing to load instructions or skills, or producing blocked tool calls—users have no visibility into why. This Skill investigates the underlying debug log files to find evidence-based root causes instead of guessing. ## Core Features & Use Cases - Log-Based Root Cause Analysis: Parses session debug logs (main.jsonl, subagent logs, system prompt and tool definition snapshots) to trace exactly what happened during a conversation. - Event Correlation: Follows span hierarchies across user messages, LLM requests, tool calls, and subagent invocations to reconstruct the full execution flow. - Cross-Platform Search Strategy: Uses grep/jq on macOS/Linux and Select-String/Node.js on Windows to efficiently search large log files without loading them into memory. - Use Case: A user asks why their custom skill did not load. The Skill searches the session log for discovery events, finds a name mismatch between the folder and SKILL.md frontmatter, and explains the fix. ## Quick Start Ask why a specific chat request was slow or why a tool, skill, or agent did not load, and point to the relevant session log directory.
