A public-safe research prototype for controlling AI-agent tool actions with deterministic policy, risk-based human approval, time-bound authorization and a tamper-evident audit chain.
A cooperative guardrail for AI agents that blocks destructive shell commands, SQL statements, or cloud operations until a human approves them out-of-band, with a tamper-evident local ledger.
Enables controlled AI-agent access to enterprise-shaped tools with a deny-by-default gated write path, human approval, dry-run execution, and append-only audit logging.
Local zero-trust permission gateway for AI agents. Enforces policy-based tool authorization, human approvals, scoped permissions, and cryptographically verifiable audit logs.
A runtime gate for coding agents. Blocks the tool calls that wreck a repo (force-push main, rm -rf, secret exfiltration, CI wipe) and lets normal build and commit work through. Machine-checked git-branch core (z3); the rest is high-precision heuristics. Tested on 3,790 real CI commands, 0 false blocks.