004/

Policy Playground

Simulate how Manyr evaluates agent actions. Configure policies, submit actions, and observe real-time decisions with full audit logging.

Agent Action

Agent Type

Tool

Proposed Action

Try an example:

Require approval for high-risk actions

Actions above risk threshold need human review

Block external data exfiltration

Prevent data transfer outside organization

Restrict filesystem access

Only allow access to approved directories

Require staging-only for DB writes

Database modifications only in staging

Enforce least-privilege tool scopes

Constrain actions to minimum permissions

Risk Tolerance30%

StrictPermissive

No audit entries yet. Evaluate an action to see results.