{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:7RTYASI4KPL5FT7N3LJUSKDA5L","short_pith_number":"pith:7RTYASI4","canonical_record":{"source":{"id":"2507.19399","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-07-25T16:06:16Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f48d8fe49efbd15a8782fe544acbfe6eb7462f600f84b7d09066ac11865a77c0","abstract_canon_sha256":"aa911751f096382698e34798ef39580c20168e135e2269e8f79decb0d4b95f94"},"schema_version":"1.0"},"canonical_sha256":"fc6780491c53d7d2cfeddad3492860eac4f4a1cc98f97ddea7f6e01112daca90","source":{"kind":"arxiv","id":"2507.19399","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.19399","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"arxiv_version","alias_value":"2507.19399v1","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.19399","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"pith_short_12","alias_value":"7RTYASI4KPL5","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"pith_short_16","alias_value":"7RTYASI4KPL5FT7N","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"pith_short_8","alias_value":"7RTYASI4","created_at":"2026-07-05T11:43:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:7RTYASI4KPL5FT7N3LJUSKDA5L","target":"record","payload":{"canonical_record":{"source":{"id":"2507.19399","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-07-25T16:06:16Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f48d8fe49efbd15a8782fe544acbfe6eb7462f600f84b7d09066ac11865a77c0","abstract_canon_sha256":"aa911751f096382698e34798ef39580c20168e135e2269e8f79decb0d4b95f94"},"schema_version":"1.0"},"canonical_sha256":"fc6780491c53d7d2cfeddad3492860eac4f4a1cc98f97ddea7f6e01112daca90","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:43:24.891289Z","signature_b64":"nlxmVDTcnAsEqWpyWyHMPEFneS9qP2u7MT9StyjWIsVV3LkWLat479+GwqLIoM1tUKRtyoY2z+XctokjSewKBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fc6780491c53d7d2cfeddad3492860eac4f4a1cc98f97ddea7f6e01112daca90","last_reissued_at":"2026-07-05T11:43:24.890845Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:43:24.890845Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2507.19399","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:43:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5P3gKgFLdV7SIJ4v5d4VlK3MoqLYfztk+zMR+x6FF3djfpxJ0p5BNcLWiuJPfnHlwXunuGyUvYcpTyZCQqgJCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T03:40:06.805362Z"},"content_sha256":"205f5b67155c011b33a5500239519d16ba3d5b4a83a9a64544a55d669c3a12b4","schema_version":"1.0","event_id":"sha256:205f5b67155c011b33a5500239519d16ba3d5b4a83a9a64544a55d669c3a12b4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:7RTYASI4KPL5FT7N3LJUSKDA5L","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Running in CIRCLE? A Simple Benchmark for LLM Code Interpreter Security","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Gabriel Chua","submitted_at":"2025-07-25T16:06:16Z","abstract_excerpt":"As large language models (LLMs) increasingly integrate native code interpreters, they enable powerful real-time execution capabilities, substantially expanding their utility. However, such integrations introduce potential system-level cybersecurity threats, fundamentally different from prompt-based vulnerabilities. To systematically evaluate these interpreter-specific risks, we propose CIRCLE (Code-Interpreter Resilience Check for LLM Exploits), a simple benchmark comprising 1,260 prompts targeting CPU, memory, and disk resource exhaustion. Each risk category includes explicitly malicious (\"di"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.19399","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/2507.19399/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:43:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RMJE9fiPKw8yuP0Bjb4xPMrtOKYhLPcINM1oKE0fXh10Ke5p9q2wKEwTa6b5kNBqHYdzfp78zSfsr7poUB87DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T03:40:06.805779Z"},"content_sha256":"0c0a5a03e88e5c18c661ffd9dce4285e8ce7a32d0bcfd6cc5d25c7653d9ee71f","schema_version":"1.0","event_id":"sha256:0c0a5a03e88e5c18c661ffd9dce4285e8ce7a32d0bcfd6cc5d25c7653d9ee71f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7RTYASI4KPL5FT7N3LJUSKDA5L/bundle.json","state_url":"https://pith.science/pith/7RTYASI4KPL5FT7N3LJUSKDA5L/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7RTYASI4KPL5FT7N3LJUSKDA5L/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-07T03:40:06Z","links":{"resolver":"https://pith.science/pith/7RTYASI4KPL5FT7N3LJUSKDA5L","bundle":"https://pith.science/pith/7RTYASI4KPL5FT7N3LJUSKDA5L/bundle.json","state":"https://pith.science/pith/7RTYASI4KPL5FT7N3LJUSKDA5L/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7RTYASI4KPL5FT7N3LJUSKDA5L/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:7RTYASI4KPL5FT7N3LJUSKDA5L","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"aa911751f096382698e34798ef39580c20168e135e2269e8f79decb0d4b95f94","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-07-25T16:06:16Z","title_canon_sha256":"f48d8fe49efbd15a8782fe544acbfe6eb7462f600f84b7d09066ac11865a77c0"},"schema_version":"1.0","source":{"id":"2507.19399","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.19399","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"arxiv_version","alias_value":"2507.19399v1","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.19399","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"pith_short_12","alias_value":"7RTYASI4KPL5","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"pith_short_16","alias_value":"7RTYASI4KPL5FT7N","created_at":"2026-07-05T11:43:24Z"},{"alias_kind":"pith_short_8","alias_value":"7RTYASI4","created_at":"2026-07-05T11:43:24Z"}],"graph_snapshots":[{"event_id":"sha256:0c0a5a03e88e5c18c661ffd9dce4285e8ce7a32d0bcfd6cc5d25c7653d9ee71f","target":"graph","created_at":"2026-07-05T11:43:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2507.19399/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As large language models (LLMs) increasingly integrate native code interpreters, they enable powerful real-time execution capabilities, substantially expanding their utility. However, such integrations introduce potential system-level cybersecurity threats, fundamentally different from prompt-based vulnerabilities. To systematically evaluate these interpreter-specific risks, we propose CIRCLE (Code-Interpreter Resilience Check for LLM Exploits), a simple benchmark comprising 1,260 prompts targeting CPU, memory, and disk resource exhaustion. Each risk category includes explicitly malicious (\"di","authors_text":"Gabriel Chua","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-07-25T16:06:16Z","title":"Running in CIRCLE? A Simple Benchmark for LLM Code Interpreter Security"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.19399","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:205f5b67155c011b33a5500239519d16ba3d5b4a83a9a64544a55d669c3a12b4","target":"record","created_at":"2026-07-05T11:43:24Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"aa911751f096382698e34798ef39580c20168e135e2269e8f79decb0d4b95f94","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-07-25T16:06:16Z","title_canon_sha256":"f48d8fe49efbd15a8782fe544acbfe6eb7462f600f84b7d09066ac11865a77c0"},"schema_version":"1.0","source":{"id":"2507.19399","kind":"arxiv","version":1}},"canonical_sha256":"fc6780491c53d7d2cfeddad3492860eac4f4a1cc98f97ddea7f6e01112daca90","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"fc6780491c53d7d2cfeddad3492860eac4f4a1cc98f97ddea7f6e01112daca90","first_computed_at":"2026-07-05T11:43:24.890845Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:43:24.890845Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"nlxmVDTcnAsEqWpyWyHMPEFneS9qP2u7MT9StyjWIsVV3LkWLat479+GwqLIoM1tUKRtyoY2z+XctokjSewKBA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:43:24.891289Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.19399","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:205f5b67155c011b33a5500239519d16ba3d5b4a83a9a64544a55d669c3a12b4","sha256:0c0a5a03e88e5c18c661ffd9dce4285e8ce7a32d0bcfd6cc5d25c7653d9ee71f"],"state_sha256":"f2636cc2a06d0c78de38d6fcb99d078b3457409bbdd8cdbcc2bafcdd68783c0d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/YrwrHHCVFBdXitlZX0VJlzAci+6VIdLHZIGYuYW6N+6vCOJTb4Cx3PuyQQhnrJdsppBco/1VoISGp9iAhJ/CQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T03:40:06.807765Z","bundle_sha256":"2b9e9f8d2a5e8b957a3662afea0d208e32196a7c77ffc3905c2a0ca1a543b59d"}}