research-package RP-2026-07-28-EVAL-THREAT

Evaluation System Threat Model

Evaluation System Threat Model

Assets

  • user files and Git history;
  • credentials and identity;
  • grader integrity and gold outcomes;
  • benchmark validity;
  • external services;
  • raw traces and potentially sensitive data;
  • operator time and money.

Trust Boundaries

  1. task author ↔ task fixture;
  2. agent ↔ sandbox/tools/network;
  3. agent outcome ↔ grader;
  4. grader ↔ gold data;
  5. runtime ↔ production credentials/services;
  6. results ↔ published capability claims.

Threats and Controls

Threat Failure Minimum control
Agent reads expected outcomes TF-006 Blind fixture export; gold outside agent-readable path
Agent edits grader TF-006/TF-009 Read-only external grader process
Task content injects instructions AF-007 Authority separation; adversarial fixture; action telemetry
Agent reaches production service AF-006 No production credentials; network deny-by-default
Agent escapes intended scope AF-006 Disposable sandbox; explicit tool permissions; timeout
Grader rewards shortcut TF-003/TF-009 Negative probes, exploit audit, state recomputation
Correct alternative rejected TF-002 Alternative-correct probes; semantic adjudication
Incomplete solution passes TF-003 Mutation/incomplete probes and regression checks
Trace exposes secrets IF-003/security Redaction policy; synthetic credentials; access controls
Long run drifts into harm AF-006 Runtime monitor, checkpoints, kill switch, rollback
Evaluator overclaims results TF-008 Scope statement, per-task results, uncertainty, replication gate

Architectural Consequence

Security and evaluation are not serial phases. Containment, gold separation, telemetry, pause, and rollback must exist before adversarial or long-horizon evaluation. Later security research can deepen controls, but it cannot be postponed behind baseline work.

Residual Risk

Local documentation fixtures are low consequence, but the architecture must not generalize that low risk to networked, cyber, browser, financial, or production-code tasks.