Proofpress builds an evidence ledger for AI-agent research: withdrawing a finding revokes its dependents
tallmetommy · x · 2026-10-07
Only Then Labs launched Proofpress, an open-source evidence ledger for research teams using AI agents, designed so findings survive across sessions and agents.
Key mechanics:
- Capture findings: bind each conclusion to its evidence, tested conditions and review decisions — failed experiments count as useful knowledge.
- Review before reuse: agents propose work; an authorized reviewer decides what becomes approved context, with approvals tied to the exact version reviewed.
- Carry changes forward: when a recorded finding is withdrawn, its dependents are withheld from fresh approved context until reassessed; unrelated work stays available.
The site includes a replayable demo of a local kernel run showing the withdraw-and-withhold flow.
More from coding & agent
- Creator's 10-minute AI skill-made video hits 170K views overnight — AlchainHust · 2026-10-07
- Organizer runs Tokyo dev forum with 5 AI agent workflows, caps attendance at 150 — soleio · 2026-10-07
- Consultant ditches custom voice agent build, deploys Grok bot with Cursor teams — alexcovo_eth · 2026-10-07
- Open-Source sopc: Manage Prompts Across Many Voice Agents Like Code — BadAthMOFO · 2026-10-07
- Benchmark: OpenAI Decisions API costs 2x more, 5-10% worse than Jev — xeophon · 2026-10-07
- Devs are cloning Photoshop in Rust and open-sourcing it — and AI-native tools need rethinking — ravisparikh · 2026-10-07