{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:M33RL3X4JURAPZF2TW5PTN4TBA","short_pith_number":"pith:M33RL3X4","schema_version":"1.0","canonical_sha256":"66f715eefc4d2207e4ba9dbaf9b793080099c6c98a2a61b4683c4578dd1fa400","source":{"kind":"arxiv","id":"2607.19449","version":1},"attestation_state":"computed","paper":{"title":"Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Aarushi Singh","submitted_at":"2026-07-21T13:23:05Z","abstract_excerpt":"Evaluation frameworks for tool-augmented LLM agents focus overwhelmingly on capability metrics or explicit tool crashes, leaving silent infrastructure failures and HTTP 200 responses with empty, null, or malformed payloads largely unaudited. We introduce a lightweight black-box auditing framework that injects four silent failure profiles across 12 production-adjacent tool stubs and classifies agent responses into three mutually exclusive behavioral classes: Honest Surrender (HSR), Fabrication (FAR), and Unfaithful Safety Refusal (USR). Evaluating two frontier and two open-source models at temp"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2607.19449","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-07-21T13:23:05Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"2b02335989a45b4e8ab8470a3d21631c269bae9a41cc25e0bcb42a0ddcdafb27","abstract_canon_sha256":"951d0aece8fefccb7c4b4ca2d73e309bffdc15aa3cb20d7dab2491d00c56309d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-23T00:23:50.328344Z","signature_b64":"HyEq/t6J/qtUmO5UzcYS7N/BTAUG51qdKmTfBLLsHFq67MKCwZLKyCF4xvWl/dqgbuzoiYA7ZCrQuA3g5OIfBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"66f715eefc4d2207e4ba9dbaf9b793080099c6c98a2a61b4683c4578dd1fa400","last_reissued_at":"2026-07-23T00:23:50.327583Z","signature_status":"signed_v1","first_computed_at":"2026-07-23T00:23:50.327583Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Aarushi Singh","submitted_at":"2026-07-21T13:23:05Z","abstract_excerpt":"Evaluation frameworks for tool-augmented LLM agents focus overwhelmingly on capability metrics or explicit tool crashes, leaving silent infrastructure failures and HTTP 200 responses with empty, null, or malformed payloads largely unaudited. We introduce a lightweight black-box auditing framework that injects four silent failure profiles across 12 production-adjacent tool stubs and classifies agent responses into three mutually exclusive behavioral classes: Honest Surrender (HSR), Fabrication (FAR), and Unfaithful Safety Refusal (USR). Evaluating two frontier and two open-source models at temp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.19449","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2607.19449/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2607.19449","created_at":"2026-07-23T00:23:50.327999+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.19449v1","created_at":"2026-07-23T00:23:50.327999+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.19449","created_at":"2026-07-23T00:23:50.327999+00:00"},{"alias_kind":"pith_short_12","alias_value":"M33RL3X4JURA","created_at":"2026-07-23T00:23:50.327999+00:00"},{"alias_kind":"pith_short_16","alias_value":"M33RL3X4JURAPZF2","created_at":"2026-07-23T00:23:50.327999+00:00"},{"alias_kind":"pith_short_8","alias_value":"M33RL3X4","created_at":"2026-07-23T00:23:50.327999+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA","json":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA.json","graph_json":"https://pith.science/api/pith-number/M33RL3X4JURAPZF2TW5PTN4TBA/graph.json","events_json":"https://pith.science/api/pith-number/M33RL3X4JURAPZF2TW5PTN4TBA/events.json","paper":"https://pith.science/paper/M33RL3X4"},"agent_actions":{"view_html":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA","download_json":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA.json","view_paper":"https://pith.science/paper/M33RL3X4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.19449&json=true","fetch_graph":"https://pith.science/api/pith-number/M33RL3X4JURAPZF2TW5PTN4TBA/graph.json","fetch_events":"https://pith.science/api/pith-number/M33RL3X4JURAPZF2TW5PTN4TBA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA/action/storage_attestation","attest_author":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA/action/author_attestation","sign_citation":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA/action/citation_signature","submit_replication":"https://pith.science/pith/M33RL3X4JURAPZF2TW5PTN4TBA/action/replication_record"}},"created_at":"2026-07-23T00:23:50.327999+00:00","updated_at":"2026-07-23T00:23:50.327999+00:00"}