Testing agent-written specs against every historical CVE, then adversarial attack
jd_pressman · x · 2026-10-24
Responding to a quip that Anthropic is rediscovering the hard lesson of formal specification, jdpressman outlines his plan: let the agent pick the spec, then walk through every historical CVE asking whether the spec would in principle have stopped it, followed by an adversarial pass in breaker mode attacking the software and spec to find where they're weak. Another commenter crowned the idea worthy of "Ambassador of the security community."
Related event: Testing Agent-Written Specs by Replaying Historical CVEs Against Them(4 posts)→
More from coding & agent
- Factory tested Reflection's Beam pre-launch via Together AI for day-0 support — togethercompute · 2026-10-24
- Benzi: compiler-backed coding agent hits 82.4% SWE-bench Verified for under $13 — DonkeyTheKing · 2026-10-24
- Bench shows GPT-5.6/6 Luna crush multi-agent coordination while DeepSeek V4 fails — JoshPurtell · 2026-10-24
- Payman AI launches MCP server for natural-language payments and payee management — modelcontextprotocol · 2026-10-24
- Grok Bot adds secure vault so agents can log in and pay without seeing your secrets — XFreeze · 2026-10-24
- Comtrade MCP server pipes UN bilateral trade data into LLMs — modelcontextprotocol · 2026-10-24