What problem does it solve? Working in unfamiliar, messy, or undocumented codebases forces developers to rediscover the same flows, gotchas, and conventions repeatedly, and AI assistants often guess or invent answers instead of verifying them. ## Core Features & Use Cases - Structured Mission Cycle: Follows a six-phase workflow (Briefing, Recon, Plan, Execute, Verify, Debrief) so every bug fix, feature, or investigation is planned and verified before changes land. - Persistent .notebook/ Knowledge Base: Captures flows, patterns, gotchas, and domain terms as pointer-based notes indexed in INDEX.md, so discoveries survive across sessions. - Knowledge Verification Chain: Verifies APIs and framework behavior against project docs, Context7 MCP, or web search instead of relying on training memory, and explicitly flags uncertainty. - Use Case: A developer reports a 500 error when applying a coupon in an unfamiliar project. The skill traces the checkout flow, finds a snake_case/camelCase schema mismatch, proposes a verified fix plan, and documents the API convention in .notebook/ for future sessions. ## Quick Start Ask the agent to fix a bug, implement a feature, or explain how a flow works in your current codebase, and it will investigate, plan, execute, and record its findings in .notebook/.
