escalation

name	escalation
description	Decide when an agent must stop and ask a human, higher-authority reviewer, or specialist agent before continuing. Use when a task involves destructive actions, credentials, production systems, external side effects, legal/security risk, ambiguous instructions, failed validation, or unresolved conflict.

Escalation

Prevent agents from crossing risk boundaries without explicit approval or specialist review.

Core procedure

1. Identify the action the agent wants to take and its potential side effects.
2. Classify risk: low, medium, high, critical using references/risk-tier-model.md.
3. Check automatic escalation triggers: secrets, production, deletion, external communication, legal/financial claims, security exposure, failed tests, unresolved conflict, unclear authority.
4. Define what information the approver needs: options, recommendation, evidence, risk, rollback, deadline.
5. Ask for approval only for the specific next action, not broad future permission.
6. Record the approval outcome and any conditions before continuing.

Output format

Produce a concise, auditable markdown artifact. Include:

## Escalation

**Scope:** ...
**Decision / recommendation:** ...
**Risk level:** low / medium / high / critical

### Findings
- ...

### Plan / matrix / record
| Item | Owner | Scope | Evidence | Gate / next action |
|---|---|---|---|---|

### Escalations
- ...

### Next actions
1. ...

Expected output focus: escalation request with action, risk, evidence, options, recommendation, approval needed, safe fallback.

Guardrails

• Prefer the narrowest capable agent and least authority required.
• Do not let an agent approve its own high-risk work.
• Mark assumptions separately from confirmed facts.
• Require human approval for destructive, credentialed, production, external, legal, financial, or security-sensitive actions.
• Do not invent agent capabilities; use the roster, plugin manifests, source files, and explicit user instructions.

Related references

• references/escalation-policy.md
• references/risk-tier-model.md
• references/authority-model.md