What problem does it solve? When a bug report or failing test names a symptom but the code path producing it is unknown, developers often guess at fixes without understanding how execution actually reaches the failure. This Skill enforces tracing the full call chain from symptom to entry point before any hypothesis or fix is proposed. ## Core Features & Use Cases - Symptom-to-code location: Find the relevant code via semantic search or baseline grep on error messages, log lines, and exception types. - Bidirectional caller/callee tracing: Walk both callers and callees outward until reaching a test or public entry point such as an HTTP handler, CLI command, or message consumer. - Recent-change detection: Use git log, git blame, or change-detection tooling to check whether a recent commit introduced the bug before assuming it is old. - Use Case: A failing test reports a wrong status code. Use this Skill to locate the code emitting it, trace callers back to the HTTP handler, follow the execution path, and identify the commit that introduced the regression before writing a fix. ## Quick Start Ask the agent to trace why this test fails by walking the call chain from the error message back to its entry point before suggesting any fix.