{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UFM7MDIAJS6LUT6WXHEMVO4UUA","short_pith_number":"pith:UFM7MDIA","schema_version":"1.0","canonical_sha256":"a159f60d004cbcba4fd6b9c8cabb94a0347c2ab9cb4db47a26d5124864f91204","source":{"kind":"arxiv","id":"2508.00271","version":2},"attestation_state":"computed","paper":{"title":"MetaAgent: Toward Self-Evolving Agent via Tool Meta-Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.IR"],"primary_cat":"cs.AI","authors_text":"Hongjin Qian, Zheng Liu","submitted_at":"2025-08-01T02:30:32Z","abstract_excerpt":"In this work, we propose MetaAgent, an agentic paradigm inspired by the principle of learning-by-doing, where expertise is developed through hands-on practice and continual self-improvement. MetaAgent starts with a minimal workflow, equipped only with basic reasoning and adaptive help-seeking abilities. When a knowledge gap is encountered, MetaAgent generates natural language help requests, which are routed to the most suitable external tool by a dedicated tool router. As MetaAgent solves tasks, it continually conducts self-reflection and answer verification, distilling actionable experience i"},"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":"2508.00271","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-01T02:30:32Z","cross_cats_sorted":["cs.CL","cs.IR"],"title_canon_sha256":"3babb436640b46ac1d752b8ccbd36052e640a0a98ef0e0aa681522bf50c641fa","abstract_canon_sha256":"9e6b6ad1a42db589dcc219fc79a013ede668094c317ae16179965240161348dc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:02:49.364564Z","signature_b64":"ggNsUm701k2o3raPXqAruA/oaYNUqY15DSZBXvZGVU8whB13NSGV5sNw3ceGXTWlrqfo6LeQKk8eUTLCtr61Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a159f60d004cbcba4fd6b9c8cabb94a0347c2ab9cb4db47a26d5124864f91204","last_reissued_at":"2026-07-05T12:02:49.364020Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:02:49.364020Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MetaAgent: Toward Self-Evolving Agent via Tool Meta-Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.IR"],"primary_cat":"cs.AI","authors_text":"Hongjin Qian, Zheng Liu","submitted_at":"2025-08-01T02:30:32Z","abstract_excerpt":"In this work, we propose MetaAgent, an agentic paradigm inspired by the principle of learning-by-doing, where expertise is developed through hands-on practice and continual self-improvement. MetaAgent starts with a minimal workflow, equipped only with basic reasoning and adaptive help-seeking abilities. When a knowledge gap is encountered, MetaAgent generates natural language help requests, which are routed to the most suitable external tool by a dedicated tool router. As MetaAgent solves tasks, it continually conducts self-reflection and answer verification, distilling actionable experience i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.00271","kind":"arxiv","version":2},"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/2508.00271/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":"2508.00271","created_at":"2026-07-05T12:02:49.364091+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.00271v2","created_at":"2026-07-05T12:02:49.364091+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.00271","created_at":"2026-07-05T12:02:49.364091+00:00"},{"alias_kind":"pith_short_12","alias_value":"UFM7MDIAJS6L","created_at":"2026-07-05T12:02:49.364091+00:00"},{"alias_kind":"pith_short_16","alias_value":"UFM7MDIAJS6LUT6W","created_at":"2026-07-05T12:02:49.364091+00:00"},{"alias_kind":"pith_short_8","alias_value":"UFM7MDIA","created_at":"2026-07-05T12:02:49.364091+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.24428","citing_title":"Escaping the Self-Confirmation Trap: An Execute-Distill-Verify Paradigm for Agentic Experience Learning","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22385","citing_title":"MetaPS: Adaptive Programmatic Strategy Selection for Market Agents","ref_index":91,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA","json":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA.json","graph_json":"https://pith.science/api/pith-number/UFM7MDIAJS6LUT6WXHEMVO4UUA/graph.json","events_json":"https://pith.science/api/pith-number/UFM7MDIAJS6LUT6WXHEMVO4UUA/events.json","paper":"https://pith.science/paper/UFM7MDIA"},"agent_actions":{"view_html":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA","download_json":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA.json","view_paper":"https://pith.science/paper/UFM7MDIA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.00271&json=true","fetch_graph":"https://pith.science/api/pith-number/UFM7MDIAJS6LUT6WXHEMVO4UUA/graph.json","fetch_events":"https://pith.science/api/pith-number/UFM7MDIAJS6LUT6WXHEMVO4UUA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA/action/storage_attestation","attest_author":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA/action/author_attestation","sign_citation":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA/action/citation_signature","submit_replication":"https://pith.science/pith/UFM7MDIAJS6LUT6WXHEMVO4UUA/action/replication_record"}},"created_at":"2026-07-05T12:02:49.364091+00:00","updated_at":"2026-07-05T12:02:49.364091+00:00"}