OpenMake
01
What changed
The current partial week strengthened orchestration reliability and frontend-backend parity, then completed a path that sends attachments beyond Cloudflare's per-request limit in chunks and analyzes scanned PDFs with no text layer. Small documents use a multi-page native OCR fallback, while large documents move into the sandbox for targeted page OCR.
- Improved routing, post-processing, agent-task completion, and frontend-backend parity.
- Added authenticated chunk uploads and their rate-limit invariant, with a 30 MB creation-time OCR gate protecting Cloudflare's response window.
- Uploaded a real 66 MB, 688-page scanned PDF to chat.openmake.cc, targeted 13 pages in four agent turns, produced a report artifact, passed 5/5 unit and 135/135 related tests, then completed PR #431 CI, merge, and production deployment.
02
How this week was reconstructed
The archive is incomplete for part of this week. Claims with a session are session-backed; implementation without a recovered transcript is described only from Git evidence.
We audited the available Claude Code main sessions, their proven child lineages, and the repository history. Session notes explain intent and investigation; Git remains the authority for code that actually landed.
The final week is still in progress. Counts and claims stop at the audit cutoff of August 4, 2026.
03
Team identity and project roles
OpenMake Team is the community identity; openmake_llm is the software project; OpenMake is the product name. Non-developer maker riskpw leads openmake_llm through vibe coding, and professional developer rocky supports its development.
Source evidence