What problem does it solve?
This Skill prevents research decisions from becoming hard to trace, non-reproducible, or missing stop-loss criteria by enforcing a structured Decision Log format with evidence, milestones, and kill rules.
Core Features & Use Cases
- Two modes (format + create): Validates and repairs existing Decision Log frontmatter/body, or generates a complete Decision Log from a one-sentence decision prompt.
- Decision traceability by design: Forces required fields plus a concrete evidence chain that must reference verifiable sources (e.g., LIT-xxxx, EXP-xxxx, meeting IDs, or DOIs).
- Falsification-driven risk control: Requires a weighted decision matrix, measurable milestone table, explicit Kill Criteria with observable thresholds and physical/experimental justification, plus trigger actions.
- Lifecycle and stakeholder alignment: Maintains append-only decision lifecycle logs and generates/validates a stakeholder communication table, and checks for dangling references.
Quick Start
Tell the AI: format my existing Decision Log note (Decision Log) and ensure it includes the weighted decision matrix, evidence chain, milestones, Kill Criteria with trigger actions, resurrection conditions, stakeholder table, and lifecycle log.