Create an agent with admin permissions

Configuration for data analysis workloads

Create the FastAPI application (requires ``agent-sre[api]`` extra).

Balanced risk policy for general use

Create or update a dataset card (README.md) for an EMK dataset. Args: repo_id: The Hugging Face Hub repository ID. descr

Create or update the dataset card (README.md) for a HF dataset. Args: repo_id: Repository ID on HF Hub. description: Dataset

Create a governance layer with default alignment rules. Returns: GovernanceLayer with standard rules configured

Create an escrow for a task. Locks credits from requester until task completion or timeout.

Create a chaos experiment.

Create a trust gate node for the graph. Use as: graph.add_node("trust_gate", checkpoint.create_gate())

Generate a Hugging Face Dataset Card for IATP datasets. Args: repo_id: Repository ID for the dataset output_path: Optional p

Configuration for API integration workloads

Create a new kernel instance.

Generate a Hugging Face Model Card README.md content. Args: info: ModelCardInfo with details. Uses defaults if not provided.

Generate a Hugging Face Model Card snippet for ATR tools. Creates a markdown snippet suitable for a Model Card README. Args: tools:

Create a pipeline from stages.

Create a governed execution plan. Each step in the generated plan is validated against ``allowed_tools`` before execution is permitte

Create a verification proof for data. Args: data: The data to create proof for verification_type: Ty

Example: Create a RAG (Retrieval-Augmented Generation) pipeline. Args: orchestrator: AgentOrchestrator instance Ret

Create a staged rollout.

Create a new saga for a session.

Create a new Shared Session.

Provision a sandbox with optional policy-driven constraints.

Create a new agent session with specified permissions. Args: agent_id: Unique identifier for the agent

Create a tool registry with standard built-in tools. Returns: ToolRegistry with common tools pre-registered

Create A2A task with trust verification. Args: peer_did: Target agent DID. message: Task message. peer_ca

Create a trust-gated tool from a function. This is a convenience function for wrapping simple functions with trust verification. Args:

Convenience function to wrap multiple nodes with trust verification. Args: identity: AgentIdentity for trust context nodes:

Create a capability manifest for an untrusted/sketchy agent.

Create a validator from YAML configuration. YAML format: name: my_principles description: Custom principles prin

Create a trust verification node. Raises TrustVerificationError if trust requirements not met.

Create a new file in the Virtual File System. Allows SDLC agents to create files in shared project state. All agents can see files c

Create a sponsorship bond between agents in a session.

UTC timestamp of key acquisition.

Manage the AGT credential vault.

Get the longest path through the graph (by duration).

Get only critical severity violations.

Get current trace context from context var.

* Public read-only view of the active DID. Caller code that previously * read `options.agentDid` and stored it should read this instead so * pos

Current hash chain value for tamper evidence.

Convenience: collect data from all successful steps.

Query a database (with governance).

Example executor for database queries In production, this would: - Execute actual SQL queries - Use connection pooling - App

Example executor for database writes In production, this would: - Execute actual write operations - Use transactions - Handl

All recorded decisions.

Decrement value atomically.

Decrement value atomically.

Decrement value atomically.

Create a default benchmark suite with standard scenarios.