Enforces characterize-before-change discipline with continuous green tests, separating structural changes from behavioral ones. Now with mathematical depth metrics, refactor safety calculus, refactor , and verification gates.
Forces data-model-first thinking before service boundaries or code, treating the schema as the geological layer everything else sits on. Now with mathematical depth metrics, architecture risk calculus, boundary integrity , and verification gates.
Probes the model's parametric knowledge boundaries using UnknownBench-style diagnostics. Detects when the model is confidently wrong about things outside its training data. Prevents epistemic hazard by flagging knowledge gaps before skills activate on voids.
Extracts reusable reasoning patterns from task outcomes, synthesizes new skills, updates CONDUCTOR selection policy, and manages skill library evolution. The compounding intelligence mechanism — the system gets smarter over time.
Enforces all verification gates (V1-V6 + ungameable components) as a standalone skill. Runs after skill orchestration to certify depth quality. The final quality gate — no answer ships without passing.
★not rated 5 3d agoA46 tokens
originalMIT
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: