stakeholder-risks-generator

To consolidate fragmented risk registers, this tool aggregates risk markers from capabilities.yaml, L3 escalation logs, and architecture notes into a single de-duplicated, severity-sorted RISKS.md file to support go/no-go decisions.

Consolidating escalation logs works by reading .autopilot-logs/escalations/*.json and ARCHITECTURE.md, extracting BLOCKED-ARCH and out-of-scope notes, scoring severities, and preserving stable risk_ids across refreshes into a product-scoped RISKS.md.

Yes, you can generate a stakeholder-facing risk register by reading risk markers like BLOCKED-EXT and DEFERRED directly from .planning/<product>/capabilities.yaml, scoring severities, and writing a governance-ready RISKS.md output.

The best way to maintain stable risk IDs is using a generator that preserves stable risk_ids across refreshes, ensuring continuity by de-duplicating aggregated risk signals and writing them consistently into .planning/PRODUCT/RISKS.md.

Yes, you need ARCHITECTURE.md and .planning/<product>/capabilities.yaml files present, as the generator reads out-of-scope notes and risk markers from these specific inputs to produce the final consolidated risk register.

Risk signals are scattered because they originate across capabilities.yaml, escalation logs, and architecture notes; this tool resolves fragmentation by aggregating and de-duplicating these inputs into one severity-sorted RISKS.md.