What problem does it solve?
Teams evaluating repositories with AgentRC often need to tailor readiness scoring to their stack and standards, but manually editing scoring logic is error-prone. This Skill guides you through picking, writing, and applying JSON policies that disable irrelevant checks, override impact levels, and set pass-rate thresholds.
Core Features & Use Cases
- Policy scaffolding: Generate a new
policies/<name>.json file with sensible defaults through a guided walkthrough of what to disable, what to raise, and which pass-rate threshold to set.
- Built-in policy examples: Start from shipped examples like
strict.json (100% pass rate), ai-only.json (AI tooling focus), or repo-health-only.json (traditional quality focus).
- Policy application and CI gating: Run readiness checks with one or more chained policies and enforce minimum maturity levels in CI using
--fail-level.
- Use Case: A platform team wants an org-wide baseline that disables observability checks for static sites while requiring a README at high impact, then layers a stricter frontend team policy on top and gates CI at maturity level 3.
Quick Start
Ask the assistant to create a new AgentRC policy named strict-frontend that disables observability, raises the readme criterion to high impact, and sets a 0.9 pass-rate threshold.
