Resume a suspended agent.

* Review code using multiple AI models

* Review code using multiple AI models

* Review code using multiple AI models

Review a pull request or diff. Args: pr_url: GitHub PR URL (will fetch diff) diff_content: Raw diff

Revoke this identity.

Revoke an identity and all its delegates.

* Recurse through the parent→child delegation graph, revoking each * reachable identity exactly once. `visited` is shared across the whole * tra

Revoke previously granted cross-session access.

Revoke all credentials for an agent.

TODO: Revoke the initiator's credentials and then attempt a handshake. Confirm that the handshake fails with a "credentials revoked" reason.

Revoke a delegation by ID.

Revoke a trust agreement by ID.

Rollback a previous execution if supported. Args: execution_id: ID of the execution to rollback

Rotate this credential, creating a new one. The old credential is marked as rotated but remains valid for a brief overlap period to a

Route ticket to appropriate specialist.

Push value to tail of list. Returns new list length.

Flowise-compatible run method. Logs input and passes through.

Run the MCP server.

Execute *command* inside the session's container.

* Run server in HTTP mode for development/testing.

* Run server in stdio mode for Claude Desktop.

* Run test scenarios (simulated for the extension)

Simulated existing agent. No changes needed here.

Run all audit benchmarks and return results.

Run all adapter benchmarks and return results.

Run all policy benchmarks and return results.

Run all Chaos Engine benchmarks and return results.

Run all Progressive Delivery benchmarks and return results.

Run all SLO benchmarks and return results.

Run all attack simulations

Run all security scans. Returns: (passed, findings) where passed=False if blocking issues found

Run all pragmatic truth tests.

Run all Trust Gateway tests.

Run all heuristic router tests.

Run all conversation manager tests.

Run all tests in the model upgrade test suite. Returns: bool: True if all tests passed, False otherwise

Run all constraint engineering tests.

Run all tests.

Run all Ghost Mode tests.

Run all test functions.

Run all tests.

Run all tests.

Run all tests

Run full compliance analysis

Run pip-audit to check for known CVEs.

Benchmark Agent-Mesh AuditLog with Merkle integrity.

Run the comparative benchmark on both approaches. Returns: (baseline_results, control_plane_results)

Benchmark credential issuance and token verification.

Run an example with decoupled execution and learning.

Benchmark hypervisor delta capture and hash chain.

Run the governed agent demo. Args: mock: If True, use mock gateway and exporter. If False, use real Azure.

Demonstrate prompt injection detection.

Demonstrate role-based capability guards.

Run the Contoso Bank governance demo.