Maclawran LLC

Aligned

xChk — accountable AI agents. Licenses an agent to a verified person who answers for it, gates each consequential action inside the agent's own harness (Claude Code, Cursor, Codex, Antigravity, Hermes), asks that person by passkey when the answer is not on file, and seals a hash-chained record that anyone can verify offline. Two papers, public benchmark.

Visit website ↗

Classification

Coverage surface: MCPEndpointAPI
Stage: Founding
Type: Open Core
Target audience: DevelopersMid-market
Deployment: Hybrid

Technical profile

Spec-grounded axes, verified by the TWG.

Interception architecture (R1): Vendor IntegrationProtocol Gateway
Policy model (R3): Deterministic
Authorization decisions (R4): DEFERALLOWDENYMODIFYSTEP_UP