What problem does it solve?
Teams accumulate exploratory research notes that never get turned into a trusted source of truth, so future readers and AI agents cannot tell what was actually decided. This Skill converts decided research into a stable canonical article while preserving the evidence trail.
Core Features & Use Cases
- Decision-confirmation gate: Blocks consolidation until the user confirms the actual decision, rejected alternatives, and rationale, preventing false certainty from being written down.
- Canonical article generation: Writes a structured article under
articles/ with Summary, Context, Decision, Rationale, Trade-offs, Alternatives considered, and Further reading sections, using status: stable frontmatter.
- Supersedes audit trail: Adds
supersedes: links in the new article and marks old research as status: deprecated with a superseded_by pointer, without deleting the original research.
- Use Case: After the team decides to adopt Yjs for CRDT-based editing, consolidate three provisional research notes into one canonical article at
articles/editor/crdt-architecture.md and deprecate the originals.
Quick Start
Ask the agent to consolidate the research notes on your chosen topic into a canonical article now that the team has made a decision.
