| name | approval |
| description | Inspect or set OmA approval posture (suggest, auto, full-auto) for autonomous workflow control. |
Run OmA approval-posture manager.
Input:
$ARGUMENTS
Protocol:
- Parse requested approval posture.
- Supported values:
suggest (default): propose actions and require explicit approval for side effectsauto: auto-approve low-risk routine actions, escalate high-risk actions for approvalfull-auto: proceed without routine confirmations, but still escalate destructive/high-risk operations
- If no value is provided, show current posture from
.omg/state/approval.json when available.
- Apply guardrails by posture:
- destructive operation checks
- secret/credential handling checks
- production-impact confirmation checks
- Read
.omg/state/session-lock.json before mutating shared approval state.
- If filesystem tools are available:
- if the current orchestration session owns the lock, write/update
.omg/state/approval.json
- otherwise write
.omg/state/sessions/[session-slug]/approval.json instead and report the ownership conflict
- Clarify boundary: this sets OmA policy posture and does not directly reconfigure Gemini CLI runtime approval flags.
Suggested approval.json shape:
{
"mode": "suggest|auto|full-auto",
"side_effect_policy": "confirm|risk-based|autonomous-with-guardrails",
"high_risk_requires_confirmation": true,
"updated_at": "",
"source": "oma:approval"
}
Response:
- Keep it concise and operator-facing.
- Include
Approval Status, Guardrail Policy, and Next Command.
