{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3P7DGRWPP53YOMH3IZ64ECM5HT","short_pith_number":"pith:3P7DGRWP","schema_version":"1.0","canonical_sha256":"dbfe3346cf7f778730fb467dc2099d3ccb400081c855839540f40dcd138268af","source":{"kind":"arxiv","id":"2605.30208","version":1},"attestation_state":"computed","paper":{"title":"Automating Low-Risk Code Review at Meta: RADAR, Risk Calibration, and Review Efficiency","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Alex Manners, Arjun Singh Banga, Audris Mockus, Brian Ellis, Chris Adams, David Molnar, Frederic Sagnes, Ging Cee Ng, Gurinder Grewal, Gursharan Singh, James Saindon, Jiefu Pei, Karthik Chengayan Sridhar, Matt Labunka, Matt Steiner, Nachiappan Nagappan, Nate Cook, Parveen Bansal, Pedro Canahuati, Peter Rigby, Prabhakar Goyal, Pratibha Udmalpet, Rujin Cao, Sean Xia, Sid Sidhu, Souvik Bhattacharya, Stacey Yan, Tianyu He, Vishal Parekh, Will Shackleton","submitted_at":"2026-05-28T16:44:07Z","abstract_excerpt":"AI-assisted coding tools have altered software production. At Meta, significant lines of code per human-landed diff grew by 105.9% year over year and per-developer diff volume rose 51%, with agentic AI responsible for over 80% of that growth. Meanwhile, the share of diffs receiving timely review has declined, exposing a widening gap between code supply and reviewer bandwidth. We ask three questions that progress from feasibility through calibration to impact: (1) can risk-stratified automation operate at scale across diverse organizations, (2) how does tuning the risk threshold affect the trad"},"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":"2605.30208","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.SE","submitted_at":"2026-05-28T16:44:07Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ec887746b82d642868442acae34a672bd7abb90a6dce14757895d07b574349d5","abstract_canon_sha256":"8f2f92628ee07a6baba90f643900aa85e0bf1f5836762f2e3c69c7d6d1374243"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T02:06:12.942726Z","signature_b64":"Ckaej30z0EcTXZS5/q4nHMKpoJVW67ZG0pXvFnJY8iHCY8kBBPP6LAz7YeTjlcpWoOjqEupYQVPLYb6uk+9hDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dbfe3346cf7f778730fb467dc2099d3ccb400081c855839540f40dcd138268af","last_reissued_at":"2026-05-29T02:06:12.942294Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T02:06:12.942294Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Automating Low-Risk Code Review at Meta: RADAR, Risk Calibration, and Review Efficiency","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SE","authors_text":"Alex Manners, Arjun Singh Banga, Audris Mockus, Brian Ellis, Chris Adams, David Molnar, Frederic Sagnes, Ging Cee Ng, Gurinder Grewal, Gursharan Singh, James Saindon, Jiefu Pei, Karthik Chengayan Sridhar, Matt Labunka, Matt Steiner, Nachiappan Nagappan, Nate Cook, Parveen Bansal, Pedro Canahuati, Peter Rigby, Prabhakar Goyal, Pratibha Udmalpet, Rujin Cao, Sean Xia, Sid Sidhu, Souvik Bhattacharya, Stacey Yan, Tianyu He, Vishal Parekh, Will Shackleton","submitted_at":"2026-05-28T16:44:07Z","abstract_excerpt":"AI-assisted coding tools have altered software production. At Meta, significant lines of code per human-landed diff grew by 105.9% year over year and per-developer diff volume rose 51%, with agentic AI responsible for over 80% of that growth. Meanwhile, the share of diffs receiving timely review has declined, exposing a widening gap between code supply and reviewer bandwidth. We ask three questions that progress from feasibility through calibration to impact: (1) can risk-stratified automation operate at scale across diverse organizations, (2) how does tuning the risk threshold affect the trad"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30208","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/2605.30208/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":"2605.30208","created_at":"2026-05-29T02:06:12.942359+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.30208v1","created_at":"2026-05-29T02:06:12.942359+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30208","created_at":"2026-05-29T02:06:12.942359+00:00"},{"alias_kind":"pith_short_12","alias_value":"3P7DGRWPP53Y","created_at":"2026-05-29T02:06:12.942359+00:00"},{"alias_kind":"pith_short_16","alias_value":"3P7DGRWPP53YOMH3","created_at":"2026-05-29T02:06:12.942359+00:00"},{"alias_kind":"pith_short_8","alias_value":"3P7DGRWP","created_at":"2026-05-29T02:06:12.942359+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/3P7DGRWPP53YOMH3IZ64ECM5HT","json":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT.json","graph_json":"https://pith.science/api/pith-number/3P7DGRWPP53YOMH3IZ64ECM5HT/graph.json","events_json":"https://pith.science/api/pith-number/3P7DGRWPP53YOMH3IZ64ECM5HT/events.json","paper":"https://pith.science/paper/3P7DGRWP"},"agent_actions":{"view_html":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT","download_json":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT.json","view_paper":"https://pith.science/paper/3P7DGRWP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.30208&json=true","fetch_graph":"https://pith.science/api/pith-number/3P7DGRWPP53YOMH3IZ64ECM5HT/graph.json","fetch_events":"https://pith.science/api/pith-number/3P7DGRWPP53YOMH3IZ64ECM5HT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT/action/storage_attestation","attest_author":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT/action/author_attestation","sign_citation":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT/action/citation_signature","submit_replication":"https://pith.science/pith/3P7DGRWPP53YOMH3IZ64ECM5HT/action/replication_record"}},"created_at":"2026-05-29T02:06:12.942359+00:00","updated_at":"2026-05-29T02:06:12.942359+00:00"}