What problem does it solve? Quantitative research teams need a consistent, auditable way to evaluate model risk before results are accepted. This Skill standardizes risk assessment by converting a ModelSpec into a validated RiskProfile with CI reports and recorded artifacts, so risk verdicts follow one canonical pipeline instead of ad-hoc judgment. ## Core Features & Use Cases - Canonical Risk Pipeline: Runs the risk-ci flow from ModelSpec or ArtifactRef through calc_risk, generate_risk_profile, and risk_verdict to produce a CI report and Artifact. - Contract and Schema Validation: Checks inputs against schemas.model.ModelSpec and outputs schemas.risk_profile.RiskProfile, returning fail or warning verdicts with explicit reasons when limits are breached. - Governed Collaboration: Records reuse decisions with organization_reuse, proposes frozen solutions via organization_solution, and uses permission and merge gates for cross-group reads and shared writes. - Use Case: A strategy group submits a new factor model; the risk group Agent validates the ModelSpec, computes risk metrics, emits a RiskProfile artifact, and returns a warning verdict listing breached thresholds for review. ## Quick Start Ask the risk group Agent to run the risk-ci flow on your ModelSpec and return the RiskProfile artifact with its verdict and reasons.
