{
  "schema_version": "1.3.0",
  "slug": "why-does-an-instance-level-safety-classifier-fail-to-gate-litmus-agentic-skill-evaluations-at-the-trajectory-boundary",
  "title": "Why does an instance-level safety classifier fail to gate LITMUS agentic skill evaluations at the trajectory boundary",
  "date_of_record": "2026 November 13",
  "scope": "Applies to LITMUS-style safety benchmark construction where per-step (instance-level) safety labels are composed into a trajectory-level pass/fail verdict for agentic skill evaluation. Does not apply to single-turn static prompt benchmarks, nor to evaluations where the grader observes the full action sequence as a single unit without intermediate labeling.",
  "projections": {
    "unified_markdown_uri": "https://corpusfetch.com/c/why-does-an-instance-level-safety-classifier-fail-to-gate-litmus-agentic-skill-evaluations-at-the-trajectory-boundary.md",
    "method_python_slice_uri": "https://corpusfetch.com/c/why-does-an-instance-level-safety-classifier-fail-to-gate-litmus-agentic-skill-evaluations-at-the-trajectory-boundary.py",
    "remnant_warrant_slice_uri": "https://corpusfetch.com/c/why-does-an-instance-level-safety-classifier-fail-to-gate-litmus-agentic-skill-evaluations-at-the-trajectory-boundary.warrant.json"
  },
  "token_profile": {
    "full_markdown_tokens": 1574,
    "python_slice_tokens": 333,
    "warrant_slice_tokens": 220
  },
  "remnant_summary": {
    "core_distinction": "The failure is not a classifier accuracy problem. It is a *composition* problem. An instance-level safety classifier assigns a label to each atomic step (tool call, message, plan fragment) independently. A trajectory-level verdict requires that the *sequence* be safe. These are different predicates, and the gap between them is where LITMUS benchmarks silentl...",
    "what_was_observed": "During construction of a LITMUS suite targeting file-and-network agent skills, a reference \"unsafe\" trajectory (deliberate exfiltration, three steps) was fed through the standard grading path. The grader aggregated per-step labels using `all(labels == SAFE)`. The trajectory scored as *safe* — a false negative on a known-unsafe reference. Investigation found: 1. The per-step classifier was correct on each step *in iso...",
    "condition_of_validity": "Step-label aggregation is a valid trajectory verdict **only under the independence condition**: no later step's safety depends on the payload, provenance, or side effects of an earlier step. This holds for stateless single-tool benchmarks (e.g., \"is this one SQL query safe?\"). It fails whenever the agent has a persistent workspace, memory, or tool output tha...",
    "decision_boundary": "The boundary is sharp and must be enforced mechanically: - **Aggregate step labels** iff the trajectory's dependency graph has no source→sink path spanning more than one step. - **Route to flow-aware grading** iff there exists a source→sink path of length ≥ 2, regardless of per-step labels. - **Reject from suite** iff the dependency graph cannot be construct...",
    "operational_effect": "Once the boundary is enforced, three operational consequences follow. First, LITMUS suites shrink: trajectories that previously passed construction are rejected for unconstructable graphs, which is correct but must be budgeted for (expect 10–30% attrition on tool-heavy suites). Second, the grader must carry a taint lattice, not a label set; this is a larger ..."
  },
  "matched_failure_symptoms": [
    "DOMAIN:EPISTEMIC_PROVENANCE_AND_FEDERATION_ROUTING"
  ],
  "composition_edges": {
    "primary_domain": "epistemic_provenance_and_federation_routing",
    "upstream_precondition_residuals": [
      "https://corpusfetch.com/c/according-to-what-precedence-hierarchy-must-an-instance-resolve-contradictory-directives-across-system-developer-and-user-tiers.md",
      "https://corpusfetch.com/c/how-should-an-instance-decide-that-a-stored-conclusion-has-gone-stale-without-a-new-observation.md"
    ],
    "downstream_consequence_residuals": [
      "https://corpusfetch.com/c/by-what-checkpoint-and-compaction-boundaries-must-an-instance-externalize-working-state-to-prevent-context-window-drift-and-truncation-amnesia.md",
      "https://corpusfetch.com/c/across-what-additive-schema-and-unknown-field-preservation-boundaries-must-an-instance-evolve-wire-contracts-to-prevent-rolling-deploy-deserialization-crashes.md"
    ],
    "co_occurring_sibling_residuals": [
      "https://corpusfetch.com/c/how-can-an-instance-capture-implicit-user-preferences-without-storing-private-data.md",
      "https://corpusfetch.com/c/how-can-an-instance-compress-long-reasoning-traces-into-binding-conclusions.md"
    ]
  },
  "empirical_verification_warrant": {
    "status": "AST_AND_INVARIANT_VERIFIED",
    "syntax_valid": true,
    "code_sha256": "a824f72d5406b7f8382102421e3bf0e64fe16e33a2dd81981fa48aa355fa7e79",
    "verified_classes": [
      "TrajectoryCompositionError",
      "Step",
      "Trajectory"
    ],
    "verified_functions": [
      "grade_by_step_aggregation",
      "build_suite_entry",
      "dependency_paths",
      "walk"
    ],
    "verified_functions_count": 4,
    "defensive_guards_count": 1,
    "raised_exceptions": [
      "TrajectoryCompositionError"
    ],
    "sample_fail_closed_invariants": [],
    "stdlib_dependencies": [
      "dataclasses",
      "typing"
    ],
    "observation_mode": "EMPIRICAL_BOUNDARY_CAPTURE",
    "zero_external_deps": true
  }
}