# CAIN-A75-024 Coding-Agent Action Gate

**Verdict:** VERIFIED

> Recomputes a coding-agent action record: authority, kind, path/command/network allowlists, approvals for privileged actions, effect consistency and claim honesty. It does not execute commands or touch the workspace.

| Session | Verdict | Actions | Blocked | Failed checks |
|---|---|---|---|---|
| 0 S-1 | ALLOWED | 3 | [] |  |

## What this proves and what it does not
- Proves: each action's kind and target are permitted, forbidden paths are refused, commands match an allowlist without shell metacharacters, network destinations are allowlisted, privileged actions have independent live approval, and an applied action is consistent with the rules.
- Does not prove: that an allowed code edit is correct or safe, that the allowlists are well chosen, or that the tool cannot bypass the gate. It is a record review.
- Verify offline: `python3 verify_offline.py run.json --pinned-key j8XmBxDZZiKkUyM0N7JRUR8m/WJcdlwX3nEuw4B4pPQ=`