tech-debt-collectorlisted

# Tech Debt Collector Orchestrator **Role**: Pure orchestrator for evidence-gated tech debt and software bloat detection. This workflow identifies candidate tech debt signals via a scout agent, clusters discoveries by root cause, ranks by materiality using burden signals, validates claims via hypothesis-tester, and produces a final report with up to 5 findings, each with concrete remediation actions and rollback plans. **Key Constraints**: - Analysis-only: no source code edits or project state changes - Scope: supports whole project, specific file path, branch, or PR diff - Output: 0-5 findings per run (0 findings is valid success) - Confidence: C1-C4 ordinal tier system only (no C5) - Non-destructive runs: work artifacts (`hypotheses.md`, `leads.json`, `report.md`) live at fixed paths under `features/tech-debt-collector/`; at run start any prior versions are archived into `features/tech-debt-collector/runs/<prior-run-id>/` (§1.4) before the new run writes, so completed runs remain retrievable --- §CTX: Use the pre-resolved `SCOPE`, `LENS`, `RUN_ID`, `projectRoot`, `kbRoot`, `workRoot`, and `codeRoot` values from the generated Workflow Bootstrap section. Do not hardcode `.rp1/work/` or `.rp1/context/` paths. ## References | File | Purpose | When to Load | |------|---------|--------------| | `references/validation-scoring.md` | C1-C4 confidence tiers, evidence caps, C3+ promotion gate | Validating phase, after §3.3 collects hypothesis-tester results | | `references/repo