{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:EOLHNYZROGPKAYAZNYAAAPIBN6","short_pith_number":"pith:EOLHNYZR","schema_version":"1.0","canonical_sha256":"239676e331719ea060196e00003d016f82e0522ee7e3604846dd9ff2c7445ab2","source":{"kind":"arxiv","id":"2602.01869","version":3},"attestation_state":"computed","paper":{"title":"Skill-Pro: Learning Reusable Skills from Experience via Non-Parametric PPO for LLM Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Skill-Pro lets LLM agents learn reusable procedural skills from past experiences without updating any parameters.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Haifeng Zhang, Haoxuan Li, Jun Wang, Mengyue Yang, Qirui Mi, Yisen Wang, Zhijian Ma","submitted_at":"2026-02-02T09:43:12Z","abstract_excerpt":"LLM-driven agents excel at sequential decision-making but often rely on on-the-fly reasoning, re-deriving solutions even in recurring scenarios. This insufficient experience reuse leads to computational redundancy and instability. To bridge this gap, we propose Skill-Pro, a framework enabling agents to autonomously learn reusable procedural skills from interaction experiences without parameter updates. By formalizing a Skill-MDP, Skill-Pro transforms passive episodic narratives into executable Skills defined by activation, execution, and termination conditions to ensure executability. To achie"},"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":"2602.01869","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-02-02T09:43:12Z","cross_cats_sorted":[],"title_canon_sha256":"9ce1cd3f78023a14abbe6a6e7c1681d1c7ddec483468097ab271327fdc2adcd2","abstract_canon_sha256":"090d34ac9a4679279642127a2cbb9daaa4c455584eeaf648ab6633dacac456bc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T02:05:40.216743Z","signature_b64":"N5CLOyQBqnqyE1lAKbgURtc6m9A//R2E+KYwVRzKNgg5Z2sDOq43VgB7H7dBM9YJceqVSoGN97ZHUo42FsqqDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"239676e331719ea060196e00003d016f82e0522ee7e3604846dd9ff2c7445ab2","last_reissued_at":"2026-05-29T02:05:40.216021Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T02:05:40.216021Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Skill-Pro: Learning Reusable Skills from Experience via Non-Parametric PPO for LLM Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Skill-Pro lets LLM agents learn reusable procedural skills from past experiences without updating any parameters.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Haifeng Zhang, Haoxuan Li, Jun Wang, Mengyue Yang, Qirui Mi, Yisen Wang, Zhijian Ma","submitted_at":"2026-02-02T09:43:12Z","abstract_excerpt":"LLM-driven agents excel at sequential decision-making but often rely on on-the-fly reasoning, re-deriving solutions even in recurring scenarios. This insufficient experience reuse leads to computational redundancy and instability. To bridge this gap, we propose Skill-Pro, a framework enabling agents to autonomously learn reusable procedural skills from interaction experiences without parameter updates. By formalizing a Skill-MDP, Skill-Pro transforms passive episodic narratives into executable Skills defined by activation, execution, and termination conditions to ensure executability. To achie"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Skill-Pro achieves superior reuse rates and significant performance gains with extreme memory compression by autonomously learning reusable procedural skills from interaction experiences without parameter updates, using Skill-MDP formalization and Non-Parametric PPO for reliable verification.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That semantic gradients combined with a PPO Gate can generate and verify high-quality skills that remain executable and non-degrading across recurring scenarios without any post-hoc exclusions or capability loss.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Skill-Pro enables LLM agents to autonomously learn, verify, and reuse compact procedural skills from experience via a Skill-MDP formalization and Non-Parametric PPO, yielding higher reuse rates, performance gains, and extreme memory compression across in-domain, cross-task, and cross-agent settings.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Skill-Pro lets LLM agents learn reusable procedural skills from past experiences without updating any parameters.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"cfdb92aac6ed8d7fea254fb87fc5088033ab64844870fb551b448d279863cb37"},"source":{"id":"2602.01869","kind":"arxiv","version":3},"verdict":{"id":"de1adf4e-d978-4769-bb94-7f76d07fb912","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T08:31:09.441739Z","strongest_claim":"Skill-Pro achieves superior reuse rates and significant performance gains with extreme memory compression by autonomously learning reusable procedural skills from interaction experiences without parameter updates, using Skill-MDP formalization and Non-Parametric PPO for reliable verification.","one_line_summary":"Skill-Pro enables LLM agents to autonomously learn, verify, and reuse compact procedural skills from experience via a Skill-MDP formalization and Non-Parametric PPO, yielding higher reuse rates, performance gains, and extreme memory compression across in-domain, cross-task, and cross-agent settings.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That semantic gradients combined with a PPO Gate can generate and verify high-quality skills that remain executable and non-degrading across recurring scenarios without any post-hoc exclusions or capability loss.","pith_extraction_headline":"Skill-Pro lets LLM agents learn reusable procedural skills from past experiences without updating any parameters."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2602.01869/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":"2602.01869","created_at":"2026-05-29T02:05:40.216113+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.01869v3","created_at":"2026-05-29T02:05:40.216113+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.01869","created_at":"2026-05-29T02:05:40.216113+00:00"},{"alias_kind":"pith_short_12","alias_value":"EOLHNYZROGPK","created_at":"2026-05-29T02:05:40.216113+00:00"},{"alias_kind":"pith_short_16","alias_value":"EOLHNYZROGPKAYAZ","created_at":"2026-05-29T02:05:40.216113+00:00"},{"alias_kind":"pith_short_8","alias_value":"EOLHNYZR","created_at":"2026-05-29T02:05:40.216113+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":26,"internal_anchor_count":26,"sample":[{"citing_arxiv_id":"2607.05202","citing_title":"EvoAgentBench: Benchmarking Agent Self-Evolution via Ability Transfer","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2606.24151","citing_title":"Metis: Bridging Text and Code Memory for Self-Evolving Agents","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2606.17929","citing_title":"PreAct: Computer-Using Agents that Get Faster on Repeated Tasks","ref_index":28,"is_internal_anchor":true},{"citing_arxiv_id":"2606.11680","citing_title":"Organize then Retrieve: Hierarchical Memory Navigation for Efficient Agents","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2606.11543","citing_title":"SkillJuror: Measuring How Agent Skill Organization Changes Runtime Behavior","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2606.03692","citing_title":"SkillPyramid: A Hierarchical Skill Consolidation Framework for Self-Evolving Agents","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22208","citing_title":"EvoIR-Agent: Self-Evolving Image Restoration Agentic System via Experience-Driven Learning","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23904","citing_title":"SkillOpt: Executive Strategy for Self-Evolving Agent Skills","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2606.29502","citing_title":"UCOB: Learning to Utilize and Evolve Agentic Skills via Credit-Aware On-Policy Bidirectional Self-Distillation","ref_index":21,"is_internal_anchor":true},{"citing_arxiv_id":"2605.30690","citing_title":"ElasticMem: Latent Memory as a Learnable Resource for LLM Agents","ref_index":29,"is_internal_anchor":true},{"citing_arxiv_id":"2606.01139","citing_title":"SkillRevise: Improving LLM-Authored Agent Skills via Trace-Conditioned Skill Revision","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23904","citing_title":"SkillOpt: Executive Strategy for Self-Evolving Agent Skills","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2605.23899","citing_title":"From Raw Experience to Skill Consumption: A Systematic Study of Model-Generated Agent Skills","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22208","citing_title":"EvoIR-Agent: Self-Evolving Image Restoration Agentic System via Experience-Driven Learning","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10923","citing_title":"Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18401","citing_title":"SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution","ref_index":37,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13037","citing_title":"MAP: A Map-then-Act Paradigm for Long-Horizon Interactive Agent Reasoning","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08386","citing_title":"SkillLens: Adaptive Multi-Granularity Skill Reuse for Cost-Efficient LLM Agents","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08526","citing_title":"Skill-CMIB: Multimodal Agent Skill for Consistent Action via Conditional Multimodal Information Bottleneck","ref_index":21,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10923","citing_title":"Dynamic Skill Lifecycle Management for Agentic Reinforcement Learning","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07358","citing_title":"A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applications","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07242","citing_title":"MEMOREPAIR: Barrier-First Cascade Repair in Agentic Memory","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2604.15097","citing_title":"From Procedural Skills to Strategy Genes: Towards Experience-Driven Test-Time Evolution","ref_index":6,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6","json":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6.json","graph_json":"https://pith.science/api/pith-number/EOLHNYZROGPKAYAZNYAAAPIBN6/graph.json","events_json":"https://pith.science/api/pith-number/EOLHNYZROGPKAYAZNYAAAPIBN6/events.json","paper":"https://pith.science/paper/EOLHNYZR"},"agent_actions":{"view_html":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6","download_json":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6.json","view_paper":"https://pith.science/paper/EOLHNYZR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.01869&json=true","fetch_graph":"https://pith.science/api/pith-number/EOLHNYZROGPKAYAZNYAAAPIBN6/graph.json","fetch_events":"https://pith.science/api/pith-number/EOLHNYZROGPKAYAZNYAAAPIBN6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6/action/storage_attestation","attest_author":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6/action/author_attestation","sign_citation":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6/action/citation_signature","submit_replication":"https://pith.science/pith/EOLHNYZROGPKAYAZNYAAAPIBN6/action/replication_record"}},"created_at":"2026-05-29T02:05:40.216113+00:00","updated_at":"2026-05-29T02:05:40.216113+00:00"}