{"paper":{"title":"Good to Go: The LOOP Skill Engine That Hits 99% Success and Slashes Token Usage by 99% via One-Shot Recording and Deterministic Replay","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"The LOOP Skill Engine records one LLM execution of a periodic agent task and converts it into a deterministic Loop Skill that replays without any further LLM calls.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chao Han, Kai Yu, Liang Wang, Xiaohua Wang, XuXiao Liang","submitted_at":"2026-05-14T01:05:35Z","abstract_excerpt":"Deploying AI agents for repetitive periodic tasks exposes a critical tension: Large Language Models (LLMs) offer unmatched flexibility in tool orchestration, yet their inherent stochasticity causes unpredictable failures, and repeated invocations incur prohibitive token costs. We present the LOOP SKILL ENGINE, a system that achieves a combined 99% success rate and 99% token reduction for periodic agent tasks through a one-shot recording, deterministic replay paradigm. On its first run, the agent executes the task with full LLM reasoning while the system transparently intercepts and records the"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"achieves a combined 99% success rate and 99% token reduction for periodic agent tasks through a one-shot recording, deterministic replay paradigm","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"that the greedy length-descending template extraction algorithm can always produce a branch-free Loop Skill that captures the task's functional intent without losing necessary conditional logic or requiring ongoing LLM reasoning","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"The LOOP Skill Engine records one LLM-powered run of a periodic task and converts it into a deterministic replay template that eliminates further LLM usage while maintaining high success rates.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"The LOOP Skill Engine records one LLM execution of a periodic agent task and converts it into a deterministic Loop Skill that replays without any further LLM calls.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"8deadffb3d8610447ec2c2db813df5e6204cf95077a3a2c53649beacd5058ff1"},"source":{"id":"2605.14237","kind":"arxiv","version":1},"verdict":{"id":"44900f3e-6767-4cb0-9f75-75cdb66d1dc9","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T02:34:36.125208Z","strongest_claim":"achieves a combined 99% success rate and 99% token reduction for periodic agent tasks through a one-shot recording, deterministic replay paradigm","one_line_summary":"The LOOP Skill Engine records one LLM-powered run of a periodic task and converts it into a deterministic replay template that eliminates further LLM usage while maintaining high success rates.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"that the greedy length-descending template extraction algorithm can always produce a branch-free Loop Skill that captures the task's functional intent without losing necessary conditional logic or requiring ongoing LLM reasoning","pith_extraction_headline":"The LOOP Skill Engine records one LLM execution of a periodic agent task and converts it into a deterministic Loop Skill that replays without any further LLM calls."},"references":{"count":15,"sample":[{"doi":"","year":2023,"title":"ReAct: Synergizing Reasoning and Acting in Language Models","work_id":"407a2351-25f1-497d-b611-f77d0292a8e6","ref_index":1,"cited_arxiv_id":"2210.03629","is_internal_anchor":true},{"doi":"","year":2023,"title":"Toolformer: Language Models Can Teach Themselves to Use Tools","work_id":"9bce40c8-cfd7-4983-80e0-c3bd4402322a","ref_index":2,"cited_arxiv_id":"2302.04761","is_internal_anchor":true},{"doi":"","year":2023,"title":"Reflexion: Language Agents with Verbal Reinforcement Learning","work_id":"778f739e-5f55-4961-8a2a-e4736a2757f4","ref_index":3,"cited_arxiv_id":"2303.11366","is_internal_anchor":true},{"doi":"","year":2023,"title":"S. Gravitas. AutoGPT: Autonomous Task Management with LLMs. GitHub, 2023","work_id":"7737a352-c232-45bc-ae3c-10dd789df5d8","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"LangChain: Building Applications with LLMs through Composability","work_id":"9bf0433e-0c75-4227-9cac-fc78fd80f267","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":15,"snapshot_sha256":"970c408e8c8dad000a4c12d0d7a5cfecbdf03654107dc630602bf008271f421d","internal_anchors":9},"formal_canon":{"evidence_count":2,"snapshot_sha256":"da4ef021f595893d99574aede55a714f348d64877ec6009dab5aabd75738881b"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}