Query with trust verification.

Query JSON data using dot notation. Args: data: JSON data to query path: Dot-notation path (e.g., "u

Return rows for *name* with timestamp >= *since*, ascending order.

Return a mock account balance.

Return a mock account balance.

Get recent episodes matching an action type.

Get all events involving a specific agent.

Get recent episodes for an agent.

Get events within a time range.

Return audit entries matching a trace ID.

Queue an event for broadcast.

Quick check if peer meets trust threshold. Uses local cache when possible for speed.

Reactivate a suspended agent.

Record a read (no enforcement).

Read from VFS.

Read line-delimited JSON receipts.

Check if the agent is ready to accept requests

Recalculate trust scores for all known agents.

Receive the next message from the internal queue. Args: timeout: Maximum seconds to wait. Returns: Message p

Create a server-side span for an incoming MCP request.

Events within the last N seconds.

Reclaim an orphaned agent by assigning a new owner and reactivating.

* Get recommended templates based on description

Record a denied capability check (possible privilege escalation).

Record a signing certificate/key renewal for a plugin.

Record failed interaction, decreasing trust.

Record the number of open incidents. Args: count: Number of currently open incidents labels: Additional labels/attrib

Persist an interaction and adjust trust accordingly.

Record an interaction between two agents.

Record an interaction outcome for trust updates.

Record an interaction outcome for trust updates.

Record the end of an operation

Record score in history.

Increment the signal-delivery counter. Args: signal_name: Signal name (e.g. ``"SIGKILL"``). agent_id: Identifier of t

Record SLO status as OTEL metrics. Args: slo_name: Name of the SLO status: Status string ("healthy", "warning", "crit

Record a task failure.

Record a tool call event. Args: session_id: Session ID. tool_name: Name of the tool called. success: Whet

Set the trust-score gauge for *agent_did*. Args: agent_did: DID of the agent. score: Current trust score value.

Record policy violations.

Redact credential-like values from a string. Args: value: String content that may contain credential-like material. Retu

Recursively redact PII fields from an object.

Compatibility alias for dictionary redaction. Args: mapping: Dictionary-like content to redact. Returns: The

Redact all nested values in a mapping. Args: mapping: A possibly nested mapping containing strings, lists, tuples

Redact detected sensitive data from a message.

Redact PII from text.

Optional model refinement. Fenced: the model only returns JSON, which the caller re validates. With no credentials configured this is a no op so t

Separate LLM call to evaluate the agent's output. Returns a score (0-1) and critique.

Evaluate the agent's response quality. Returns a score (0-1) and critique.

Force a tree refresh.

Force a refresh of all SLO data.

* Best-effort regex tokenizer used when no AST parser is available. * * This intentionally avoids pulling in a heavy SQL grammar dependency. It is

* Register the agent in the registry. Idempotent: a 409 response * (already registered) resolves successfully without throwing. * * @param di

Register a new agent. Args: entry: Agent registry entry Raises: ValueError: If agent is already reg

Register the four governance hooks with CrewAI. Returns ------- GovernanceHooks Self, for chaining. Rais

Register a custom experiment template.

* Publish this agent in the registry and upload an X3DH pre-key bundle. * * - Generates a signed pre-key and `oneTimePrekeyCount` one-time *

Register a new agent and issue credentials. Args: request: Registration request Returns: Registrati

Register a capability with the agent.

Register a context router (for caas integration)

Add a custom promotion gate. Args: gate: The gate definition to register.

Register a server-side handler for an RPC topic. Args: topic: RPC method / topic name. handler: Async callable that p