$ loading_
Add a fail-closed local policy gate to AI agent actions.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "com.decionis/mcp" MCP server from askskill: Run: claude mcp add 'com-decionis-mcp' -- npx -y @decionis/mcp
Developers can use it to locally evaluate agent actions before an AI agent calls external tools. If evaluation fails or violates policy, the action is denied by default.
DevOps or platform teams can use native pre-tool hooks to insert policy checks before every tool execution. This helps stop high-risk actions before they actually run.
When a team wants less dependence on external evaluation services, it can use the tool’s local evaluation capability to control agent behavior. It suits automation environments that require stronger deny-by-default policies.
This is an MCP tool that acts as a policy gate for AI agent actions. It emphasizes fail-closed behavior and supports local evaluation plus native pre-tool hooks.
Based on the description, it evaluates actions before tool calls via native pre-tool hooks. If evaluation fails or does not pass policy, the action should be denied.
The provided material does not include installation steps, runtime requirements, or key requirements. See the source repository for details.
Enforce permissions, approvals, sanitization, and audit for AI agent MCP calls.
Add policy enforcement, human approval, and audit logs to AI agent tool calls.
Govern agent decisions with auditable evidence, confidence calibration, and policy-based handoffs.
Gate MCP agent actions through compliance policies before execution.
Provides compliance guardrails for AI agents to operate safely within regulations.
Secure AI agents locally with cost controls, injection blocking, and action approvals.