Adding a Chain of Audit to Coding Agents
Funky_Chicken_22 · reddit · 2026-07-18
Starting from world-model-mcp, the author initially built a temporal knowledge graph and a constraint enforcement layer for coding agents. However, real-world projects revealed a more critical gap: verifiable decision history.
- Memory and rule enforcement alone are insufficient. Even if an agent makes a "correct decision," it's incredibly hard to independently prove what context it relied on or which rules it followed.
- To solve this, the author built Etch on top: it writes every tool call into a signed Merkle chain using a hybrid post-quantum signature.
- This allows history to be verified offline without relying on model vendors, providing both an enforcement layer and a cryptographic audit layer.
- They have signed over 16.3万 events so far, but adoption remains slow. While memory and context pain points are immediately obvious, the need for auditing usually only becomes urgent during reviews or compliance checks.
Related event: Approval and Audit Debates for Production AI Agents(6 posts)→
More from coding & agent
- Vovy launches a cursor-following voice agent for Mac, free tier gives 100 credits/mo — Scobleizer · 2026-10-27
- Redditor seeks real-world stacks for a persistent-memory personal AI that answers texts and emails — BrooksomidusGG · 2026-10-27
- Practical tip: use max effort for planning, xhigh for implementation in Claude Code — majidmanzarpour · 2026-10-27
- Chrome team explains WebMCP: agents get tools from your website, not a replacement for MCP — rseroter · 2026-10-27
- Agent-Native apps: why agents bypassing your UI can break business rules — rseroter · 2026-10-27
- 'My intelligence stack': developer splits AI coding work across Opus 5.5, Fable and Astra with human review gates — sh_reya · 2026-10-27