$ loading_
Adds deterministic pre-payment safety checks for AI agent transactions
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "governor" yet — see the docs or source repo.
Use governor to review this pending AI agent payment: send $1,200 to a new recipient address for an “urgent purchase,” and return a decision to approve, escalate for human review, or block, with reasons.
A deterministic risk decision explaining triggered rules, behavioral anomalies, or intent incoherence, plus the recommended action.
Use governor to determine whether this payment should be escalated for human review: the same agent initiated five small transfers to different accounts within 10 minutes for “testing the vendor settlement flow.”
A decision on whether to escalate, including behavioral anomaly analysis and a risk-level explanation.
Use governor to review this routine payment: pay a verified vendor $300 for a monthly subscription, with amount, recipient, and purpose matching historical records; explain whether it should be approved.
An executable approval decision stating that rules passed, behavior is normal, and intent is coherent.
Govern agent decisions with auditable evidence, confidence calibration, and policy-based handoffs.
Enforce permissions, approvals, sanitization, and audit for AI agent MCP calls.
Provides compliance guardrails for AI agents to operate safely within regulations.
Gate MCP agent actions through compliance policies before execution.
Run AI safety screening, risk classification, and red teaming controls.
Standardize Claude Code output, trim context, and monitor drift and tool noise.