code-investigatelisted

# Root Cause Investigator This command invokes the **bug-investigator** sub-agent for systematic issue analysis of the provided problem statement. Before invoking the sub-agent or registering artifacts, derive `EFFECTIVE_ISSUE_ID`: - If `ISSUE_ID` is non-empty, use it unchanged. - Otherwise derive a deterministic lowercase kebab-case slug from `PROBLEM_STATEMENT`. - Collapse repeated separators, trim leading/trailing `-`, and if the slug would be empty use `investigation`. Use `EFFECTIVE_ISSUE_ID` consistently for workspace paths, report paths, and agent inputs. **First emit**: Generate `RUN_ID` as a UUID. Derive `RUN_NAME` from the problem statement: a brief summary (max 60 chars) prefixed with `"Investigate: "`. On session start, emit the status change: ```bash rp1 agent-tools emit \ --workflow code-investigate \ --type status_change \ --run-id {RUN_ID} \ --name "Investigate: {brief summary}" \ --step investigating \ --data '{"status": "running"}' ``` ## STATE-MACHINE ```mermaid stateDiagram-v2 [*] --> investigating investigating --> [*] : done ``` Invoke the bug-investigator agent with the freeform problem statement and derived issue id: {% dispatch_agent "rp1-dev:bug-investigator" %} PROBLEM_STATEMENT={PROBLEM_STATEMENT}, ISSUE_ID={EFFECTIVE_ISSUE_ID}, KB_ROOT={kbRoot}, WORK_ROOT={workRoot} {% enddispatch_agent %} The agent will: - Analyze problem statement and gather context - Form and test hypotheses systematically - Add temporary debugging