{
  "schema_version": 1,
  "builder_plan": {
    "scenario": "An agent reads external documents or tool responses that may contain malicious instructions.",
    "change": "Replay attacks and legitimate lookalikes in an isolated copy of your agent workflow.",
    "baseline": "Your existing checks with the same tasks, permissions, and tool responses.",
    "steps": [
      "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.",
      "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.",
      "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.",
      "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test."
    ],
    "measure": "Successful attacks missed, legitimate actions blocked, detection delay, and review cost.",
    "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation."
  },
  "builder_review": null,
  "arxiv_id": "2609.14987",
  "title": "ActGuard: Pre-execution Action Auditing against Indirect Prompt Injection in LLM Agents",
  "summary": "ActGuard replaces coarse prompt-hardening and content-filtering defenses with a per-step audit that predicts a local tool prior, flags deviations in tool choice and parameters, and masks only evidence-verified malicious spans before regenerating the action. Its cost is additional inference at every agent step: a tool-prior prediction, contrastive analysis, and a verifier pass, which increases latency and API spend roughly proportional to trajectory length. It can fail when the injected instruction produces an action consistent with the tool prior (parameter-only manipulation that evades localization), when the verifier over- or under-masks evidence, or when legitimate but unusual tool sequences are treated as deviations.",
  "rationale": "For a small team deploying tool-using agents via APIs, this is implementable without model training or extra hardware, but its security-utility trade-off must be validated on the team's own tools and attack surface before production adoption.",
  "technique": "Pre-execution action auditing against indirect prompt injection",
  "family": "seguranca_e_guardrails",
  "recommendation": "testar",
  "published": "2026-09-14",
  "as_of": "2026-10-18",
  "editorial_status": "indexed",
  "source_url": "https://arxiv.org/abs/2609.14987",
  "independent_implementations": 2,
  "total_implementations": 2,
  "stars": 3,
  "citations": 0,
  "claims": [
    {
      "claim_id": "claim-01",
      "statement": "ActGuard replaces coarse prompt-hardening and content-filtering defenses with a per-step audit that predicts a local tool prior, flags deviations in tool choice and parameters, and masks only evidence-verified malicious spans before regenerating the action. Its cost is additional inference at every agent step: a tool-prior prediction, contrastive analysis, and a verifier pass, which increases latency and API spend roughly proportional to trajectory length. It can fail when the injected instruction produces an action consistent with the tool prior (parameter-only manipulation that evades localization), when the verifier over- or under-masks evidence, or when legitimate but unusual tool sequences are treated as deviations.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2609.14987",
      "source_page": null,
      "source_excerpt": ""
    }
  ],
  "exposure_map": [
    {
      "dimension": "quality",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "compute",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "latency",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "operations",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "compatibility",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "security",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "data_and_training",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "reproducibility",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    }
  ],
  "risks": [],
  "minimum_test": [],
  "open_questions": [],
  "independent_tests": [],
  "lab_reproductions": [],
  "report_available": false,
  "validation_tier": "unknown",
  "evidence_tier": "unknown",
  "infrastructure_basis": "unknown",
  "equations": [],
  "equations_status": "not_formula",
  "core_kind": "algorithm",
  "equations_fetched_at": "2026-10-03",
  "equations_source_url": "",
  "equations_source_sha256": "",
  "exposures_reviewed_at": "",
  "exposures_source_sha256": ""
}