FrictionMap is a local-first operational diagnostic for founder-led businesses. It turns messy workflow notes, process documentation, meeting notes, and founder pain points into an evidence-gated friction map.
The method:
- Normalizes operational evidence into cited anchors.
- Extracts signals across decision latency, delegation gaps, communication drag, process ambiguity, tool/system fragmentation, ownership confusion, reporting blindness, and automation opportunity.
- Applies a deterministic, versioned rubric for severity, leverage, and ease of intervention.
- Blocks unsupported, contradictory, tied, near-tied, or low-information findings instead of manufacturing a winner.
- Produces a reviewable root-cause diagnosis, top repair priorities, and a 30-day operating repair brief only after human approval.
The app is intentionally local-first and uses sanitized sample evidence. No client data leaves the browser and no runtime LLM is used to assign scores or override evidence. Codex accelerated the implementation by helping build and refactor the React/TypeScript domain engine, add the evidence schema and audit states, harden prompt-injection and noise handling, and create the deterministic regression suite. GPT-5.6 and Codex were used throughout the build workflow; the product boundary keeps diagnosis owned by the explicit FFC method rather than treating model output as authority.
Run locally with npm install, npm run dev, npm run test, npm run lint, and npm run build.
Built With
- codex
- gpt-5.6
- react
- typescript
- vite
Log in or sign up for Devpost to join the conversation.