What problem does it solve?
It prevents teams from wasting time aligning on requirements scattered across PRDs, RFCs, wiki exports, and pasted snippets by normalizing them into a traced, repo-grounded intake artifact.
Core Features & Use Cases
- Source inventory with roles: Captures every provided path/URL/paste and tags each item with a requirement role (product, architecture, api, data_flow, design, testing, security, unknown).
- Repo-first reconciliation and traceability: Loads
.ohc/PROJECT.md, compares claims to the codebase, and records conflicts, stale sections, and gaps without hiding uncertainty.
- Canonical indexing for execution: Writes/updates
.ohc/doc-sources.md using assets/doc-sources.template.md, creating a single index that agents use to plan and implement.
- Plan handoff with a BUILD gate: Produces a narrowed writing-plans handoff that includes unresolved questions and conflicts, then blocks implementation until the user confirms.
Quick Start
Ask your agent to run document-intake for an external PRD set of links so it produces .ohc/doc-sources.md and a plan covering unresolved conflicts.