investigate

# Investigate See the [fixed recurrence and root-cause policy](../../shared/native-skill-contract.md#recurrence-and-root-cause-policy) for repeated-obstacle handling; it governs symptom patching, escalation, and reset evidence. Diagnosis-first loop for unclear failures: failing tests, tracebacks, regressions, surprising runtime behavior. Produce root-cause claim with evidence, falsification, rejected alternatives before any fix. Use `investigate` until root cause established; then hand off to `develop` or `code-remediate`. ## Input Schema ```json { "symptom": "required failing command, traceback, runtime bug, CI failure, flaky behavior, or tool anomaly", "scope": "optional path/module/tool/CI run", "pace": "fast|full", "done_when": "one root cause is confirmed or the remaining uncertainty is explicit" } ``` ## Workflow ### 01: Create run directory Run `python PLUGIN_ROOT/shared/create_run.py --skill investigate` once. Retain its single printed path as `<run-directory>` and substitute that literal path into every later artifact path and helper argument. Never store or reuse the path through a shell variable; shell variables do not persist across tool calls. ### 02: Capture symptom and reproduction context Write `<run-directory>/symptom.md` with: - failing command or observed behavior - expected behavior - local vs CI vs external context - first known bad time or commit, if known - whether the failure is deterministic, flaky, or unknown ### 03: Gather signals...