Enforce governance policies and audit trails on AI agent tool calls.
npx skills add https://github.com/mmahalwy/cooper --skill agent-governance-mmahalwy
Please help me install this Agent Skill. Skill: agent-governance Source: https://github.com/mmahalwy/cooper/tree/main/.agents/skills/agent-governance Command: npx skills add https://github.com/mmahalwy/cooper --skill agent-governance-mmahalwy
AI agents that call external tools can perform unsafe, unauthorized, or non-compliant actions; this Skill provides patterns to enforce policy, detect dangerous intent before execution, limit risky operations, and maintain immutable audit trails so agents act within defined boundaries.
Use the agent-governance skill to wrap your agent's tool functions with policy checks, pre-flight intent classification, and append-only auditing so all tool calls are validated and recorded.
High-intent search queries and answers about installing and using this skill.
You can enforce safety policies on AI agents by applying composable governance decorators to tool functions, implementing allowlists, blocklists, and rate limits to restrict risky operations before execution.
Pre-flight intent classification for AI agent security evaluates tool call prompts using weighted threat signals to catch prompt injection, data exfiltration, and privilege escalation attempts before any tool is invoked.
You can maintain an audit trail for AI agent tool calls by wrapping tool functions with governance decorators that automatically emit append-only audit entries, recording every validated action for compliance review.
Yes, you can require human approval for sensitive actions by configuring composable governance policies with human-approval hooks, ensuring that restricted operations pause for review before execution.
You can prevent prompt injection attacks in multi-agent systems by applying pre-flight intent classification to detect malicious inputs and tracking inter-agent trust with decay-based scoring to limit delegations.
No, you do not need external dependencies to add governance controls; this skill provides self-contained patterns for policy enforcement, intent classification, trust scoring, and auditing without external libraries.
GabeujinImplement policy enforcement, intent classification, trust scoring, and audit trails for AI agent systems.
IgorGanapolskyBlock unsafe or non-compliant crypto trades before execution.
ZESCODEIntercept and validate system commands and file operations against safety patterns.
alsk1992Manage command approvals, tool access policies, and security modes via CLI and TypeScript API.
