Operations Advanced 2 minute read Updated 2026-06-26 UTC

Incident response

Detection, containment, investigation, recovery, and learning for model, router, evaluator, registry, and policy incidents.

Research statusEstablished incident-response principles adapted to model systems Publication statePublished Reviewed byMichael Kappel Source reports3

Model incidents cross layers

An incident may originate in data, model behavior, routing, tools, package integrity, evaluator logic, or policy. Response must preserve evidence while rapidly reducing user and system harm.

Severity examples

  • Critical: unauthorized tool or network action, sensitive-data exfiltration, control-plane compromise, failed emergency stop.
  • High: hard safety invariant violation, poisoned package in canary, widespread critical-slice regression.
  • Medium: material quality or calibration regression, repeated route failure, cost runaway within containment.
  • Low: isolated timeout, stale evidence, non-critical documentation gap.

Response workflow

pseudocode
PROCEDURE respond_to_incident(signal)
    incident <- OPEN_INCIDENT(signal, NOW_UTC())
    FREEZE_RELEVANT_ALIASES(incident.scope)
    REDUCE_OR_STOP_TRAFFIC(incident.scope)
    REVOKE_CREDENTIALS_AND_SIGNERS_IF_NEEDED(incident)
    PRESERVE_LOGS_PACKAGES_AND_POLICY_STATE(incident)

    affected <- TRACE_LINEAGE_AND_REQUESTS(incident)
    recovery <- SELECT_VERIFIED_RECOVERY_TARGET(affected)
    EXECUTE_RECOVERY(recovery)
    VERIFY_SERVICE_AND_SAFETY(recovery)

    root_cause <- INVESTIGATE_WITH_SEPARATE_TEAM(incident)
    CREATE_CORRECTIVE_BEADS(root_cause)
    EXPAND_EVALUATION_AND_CONTROLS(root_cause)
    CLOSE_ONLY_AFTER_EVIDENCE_REVIEW(incident)
END PROCEDURE

Freeze before breeding

During an incident, suspend automated candidate generation and structural changes unless they are part of an approved recovery plan. Otherwise the population can change while investigators are trying to reconstruct the event.

Evidence preservation

Capture exact artifacts, aliases, router and prompt versions, evaluator versions, policy, credentials, request traces, tool calls, and resource state. Store hashes and timestamps in UTC. Avoid modifying the affected packages during analysis.

Recovery targets

The prior champion is not automatically safe. The incident may affect a shared ancestor, dataset, runtime, or policy. Verify the recovery target against the current threat before rollback.

Post-incident learning

Add deterministic controls where possible, expand hidden suites, correct contracts, improve observability, and revise runbooks. Training a new model without fixing the control failure is incomplete remediation.

Exercises

Run tabletop and staging exercises for compromised signer, poisoned federated round, evaluator leakage, runaway cost, model exfiltration attempt, and rollback failure. Measure detection and recovery time.

Source reports used for this guide

These reports are preserved verbatim in the site archive. The guide above is an editorial synthesis and may narrow, qualify, or reorganize claims from the source material.