Check whether this backend is available and operational. Returns: True if the backend can accept evaluate() calls.

Kubernetes-style liveness probe.

Agent that does work in small steps.

Get hash field value.

Get hash field value.

Get hash field value.

Get all hash fields.

Get all hash fields.

Get all hash fields.

Get only high severity violations.

Show trust score history for an agent over time. AGENT_ID is the DID of the agent (e.g. did:mesh:agent-alpha-001).

Calculate overall cache hit rate.

Get all hash field names.

Get all hash field names.

Get all hash field names.

Human-readable label for this hooks instance.

Set hash field value.

Set hash field value.

Set hash field value.

A plain Hypervisor with no integration adapters.

Assistant ID

Create a test agent identity.

Create a test AgentIdentity with private key.

Get this agent identity.

The agent's identity.

* Ed25519 public key used to sign pre-keys. Distinct from * `identityKey.publicKey` (X25519). Peers MUST be given this key to * verify the signa

Create a second test AgentIdentity.

Import an OpenAI function definition as a Semantic Kernel plugin. Args: plugin_name: Name to register the plugin under.

Import reputation data from other nodes. This enables network-wide reputation propagation.

Increment value atomically.

Increment value atomically.

Increment value atomically.

Increment value by amount.

Increment value by amount.

* Increment the violation counter by one. * * Updates the display immediately using the existing breakdown.

Same shape, drawn from the new index-based output.

Agent that never returns (for kill testing).

Ingest a document for processing. The service will: 1. Process the raw content 2. Auto-detect the document type and structure

Ingest a reliability signal.

Initialize a new governed agent in 30 seconds. Creates the scaffolding for a governed agent with identity, trust, and audit built in.

Initialize AgentMesh integration with existing tools. Examples: # Setup Claude Desktop to use AgentMesh proxy agentmesh ini

Initialize the trust middleware.

Initialize all registered plugins

Initialize a new task step.

Mark a parameter for dependency injection. Use this as a default value for parameters that should be injected by the runtime when the tool is

Inject a fault into a running experiment.

Inject current trace context into a message. Args: message: Message to inject trace into Returns: Message w

Show detailed trust information for a specific agent. AGENT_ID is the DID of the agent to inspect (e.g. did:mesh:agent-alpha-001).

Inspect all selected stdio MCP servers in a config file.

Install a plugin from the registry.

Install a plugin from the registry.

Create a concrete experiment from this template.

Verify or generate module integrity manifest.

Intercept a tool call and return allow/deny decision.

Create an invalid manifest file for testing.

View the agent inventory.

Invoke a tool with trust verification. Args: tool: Tool name or TrustGatedTool instance input_data: Input to pass to

Governed synchronous invocation. Args: input_data: Input to pass to the chain/agent. **kwargs

Delegate to LlamaIndex ``.retrieve()`` method.

Execute the underlying query and return source nodes for scanning.

Synchronous wrapper around :meth:`invoke`. Runs the async ``invoke`` in a new event loop via ``asyncio.run()``. Useful for scripts o