AgenticContract

MCP Tools

Complete reference for AgenticContract MCP tools

AgenticContract exposes 22 core tools and 16 advanced tools via the Model Context Protocol.

Contract Tools

`contract_create`

Create a new contract between agents or agent and user.

Parameter	Type	Required	Description
`name`	string	Yes	Contract name
`parties`	string[]	Yes	List of party identifiers
`description`	string	No	Contract description

Returns: Contract ID and creation timestamp.

`contract_sign`

Sign a contract to indicate acceptance.

Parameter	Type	Required	Description
`contract_id`	string	Yes	Contract to sign
`signer`	string	Yes	Identity of the signer

Returns: Signature confirmation.

`contract_verify`

Verify a contract's validity and signature chain.

Parameter	Type	Required	Description
`contract_id`	string	Yes	Contract to verify

Returns: Validity status and signature details.

`contract_list`

List contracts with optional status filter.

Parameter	Type	Required	Description
`status`	string	No	Filter: active, expired, all

Returns: Array of contract summaries.

`contract_get`

Get a specific contract by ID.

Parameter	Type	Required	Description
`contract_id`	string	Yes	Contract ID

Returns: Full contract details.

Policy Tools

`policy_add`

Add a policy rule governing agent behavior.

Parameter	Type	Required	Description
`label`	string	Yes	Human-readable policy name
`scope`	string	Yes	`global`, `session`, or `agent`
`action`	string	Yes	`allow`, `deny`, `require_approval`, or `audit_only`
`description`	string	No	Detailed policy description
`tags`	string[]	No	Categorization tags
`conditions`	string[]	No	Prerequisite condition labels

Returns: Policy ID.

`policy_check`

Check if an action is allowed under current policies.

Parameter	Type	Required	Description
`action`	string	Yes	Description of the intended action
`scope`	string	No	Filter to specific scope

Returns: Policy action (allow, deny, require_approval, audit_only) and matching policy details.

`policy_list`

List active policies with optional scope filter.

Parameter	Type	Required	Description
`scope`	string	No	Filter: `global`, `session`, `agent`

Returns: Array of policy objects with ID, label, scope, action, and tags.

Risk Limit Tools

`risk_limit_set`

Set a risk limit threshold for a resource or action.

Parameter	Type	Required	Description
`label`	string	Yes	Limit name
`max_value`	number	Yes	Maximum allowed value
`limit_type`	string	Yes	`rate`, `threshold`, `budget`, or `count`
`window_secs`	number	No	Time window in seconds (for rate limits)

Returns: Risk limit ID.

`risk_limit_check`

Check if an action would exceed risk limits.

Parameter	Type	Required	Description
`label`	string	Yes	Limit to check against
`amount`	number	Yes	Proposed usage amount

Returns: Whether the limit would be exceeded, current usage, remaining capacity.

`risk_limit_list`

List all risk limits with current values.

Returns: Array of risk limits with ID, label, type, current value, max value, and usage ratio.

Approval Tools

`approval_request`

Request approval for a controlled action.

Parameter	Type	Required	Description
`rule_id`	string	Yes	Approval rule that triggered this
`action_description`	string	Yes	What the agent wants to do
`requestor`	string	Yes	Who is requesting

Returns: Approval request ID with pending status.

`approval_decide`

Approve or deny a pending approval request.

Parameter	Type	Required	Description
`request_id`	string	Yes	Request to decide
`decision`	string	Yes	`approve` or `deny`
`decider`	string	Yes	Who is deciding
`reason`	string	Yes	Justification

Returns: Decision ID and updated request status.

`approval_list`

List approval requests with optional status filter.

Parameter	Type	Required	Description
`status`	string	No	Filter: `pending`, `approved`, `denied`, `expired`

Returns: Array of approval requests.

Condition Tools

`condition_add`

Add a conditional execution rule.

Parameter	Type	Required	Description
`label`	string	Yes	Condition name
`condition_type`	string	Yes	`threshold`, `time_based`, `dependency`, or `custom`
`expression`	string	Yes	Condition expression

Returns: Condition ID.

`condition_evaluate`

Evaluate whether conditions are met for an action.

Parameter	Type	Required	Description
`condition_id`	string	Yes	Condition to evaluate

Returns: Whether the condition is met, with status and last evaluation result.

Obligation Tools

`obligation_add`

Add an obligation that an agent must fulfill.

Parameter	Type	Required	Description
`label`	string	Yes	Obligation name
`description`	string	Yes	Detailed description
`assignee`	string	Yes	Responsible agent/user
`deadline`	string	No	ISO 8601 deadline

Returns: Obligation ID.

`obligation_check`

Check the status of obligations.

Parameter	Type	Required	Description
`assignee`	string	No	Filter by assignee
`status`	string	No	Filter: `pending`, `fulfilled`, `overdue`, `waived`

Returns: Array of obligations, with overdue items flagged.

Violation Tools

`violation_report`

Report a contract or policy violation.

Parameter	Type	Required	Description
`description`	string	Yes	What happened
`severity`	string	Yes	`info`, `warning`, `critical`, or `fatal`
`actor`	string	Yes	Who committed the violation
`policy_id`	string	No	Related policy ID
`context`	object	No	Additional JSON context

Returns: Violation ID with timestamp.

`violation_list`

List recorded violations with optional severity filter.

Parameter	Type	Required	Description
`severity`	string	No	Minimum severity filter
`actor`	string	No	Filter by actor

Returns: Array of violation records.

Context and Stats Tools

`contract_context_log`

Log the intent and context behind a contract action.

Parameter	Type	Required	Description
`intent`	string	Yes	Why the action is being taken
`topic`	string	No	Category (e.g., "compliance", "risk")
`decision`	string	No	What was concluded

Returns: Confirmation of logged context.

`contract_stats`

Get summary statistics for the contract store.

Returns: Counts for all entity types, file size, last modified timestamp.

Error Handling

All tool errors are returned as MCP content with isError: true:

{
  "content": [{"type": "text", "text": "Error: Policy not found"}],
  "isError": true
}

Unknown tools return JSON-RPC error code -32803 (TOOL_NOT_FOUND).