Delimit Gov Impl
_delimit_gov_implDispatch any governance action—health, status, policy, evaluate, new_task, run, or verify—from one entry point, with license gating and deterministic error responses.
Instructions
Unified governance entry point — dispatches to one of seven actions.
When to use: as the single MCP-registered governance surface (delimit_gov) when the caller wants to pick the action by name in one call rather than choosing a specific delimit_gov_* alias. When NOT to use: from internal code paths — prefer the specific alias (delimit_gov_health, delimit_gov_evaluate, etc.) for clarity and so docstrings and license gates show up at the right call site.
Sibling contrast: each delimit_gov_ wrapper above is a thin alias over this implementation; they exist so the action's docstring lives at the right name. This is the dispatch core.
Side effects: action="health" / "status" are read-only and not
gated. action="policy" / "evaluate" / "new_task" / "run" / "verify"
are gated by require_premium — unlicensed callers receive a
license payload and no backend call is made. Each gated action
routes to a distinct backends.governance_bridge function (health,
status, policy, evaluate_trigger, new_task, run_task, verify) and
the result is wrapped via _with_next_steps for orchestrator hints.
Errors are deterministic ({"error": ...}); inputs that cannot be
coerced (e.g. malformed context for evaluate) short-circuit
before the backend call.
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| repo | No | Repository path. Default "." (cwd). | . |
| scope | No | Task scope (used only when action="new_task"). Required for new_task. | |
| title | No | Task title (used only when action="new_task"). Required for new_task. | |
| action | No | Which governance operation to perform. One of "health", "status", "policy", "evaluate", "new_task", "run", "verify". Default "health". Other values return a deterministic error. | health |
| context | No | Additional context (used only when action="evaluate"). Strings are auto-coerced to {"text": ...} via _coerce_dict_arg; dicts are passed through. None is allowed. | |
| task_id | No | Task ID (used only when action="run" or action="verify"). Required for those actions. | |
| risk_level | No | Risk level low/medium/high/critical (used only when action="new_task"). Default "medium". | medium |
| eval_action | No | The proposed action name to evaluate (used only when action="evaluate"). Empty string is rejected by the backend. |
Output Schema
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||