{ "schema_version": 1, "builder_plan": { "scenario": "A product assistant retrieves plausible documents but gives unsupported answers.", "change": "Change one retrieval or ranking step while keeping the document snapshot fixed.", "baseline": "Your current retrieval pipeline with identical questions and answer model.", "steps": [ "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.", "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.", "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.", "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test." ], "measure": "Supported answers, retrieval misses, correct abstentions, index cost, and response latency.", "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation." }, "builder_review": null, "arxiv_id": "2609.37673", "title": "KUPAS MASTER: Distilling the Tacit Expertise of Master Practitioners into Agent-Ready Experience Corpora", "summary": "It replaces naive retrieval over raw work records with a curated experience corpus (rules, constraints, negative examples, corner cases, callable skills) built from practitioner records and interviews through a nine-layer extraction process. The cost is primarily human and engineering effort: interviews, semantic alignment, cross-review, and ongoing consolidation of 1,576 source files into tens of thousands of curated assets, rather than GPU or model-training spend. It can fail when expert time is unavailable, when domains drift so curated rules and boundaries go stale, or when the reported gains do not transfer since evaluation is internal, sample-based (20 practitioners), and lacks independent replication or public tooling.", "rationale": "The quality gain over raw RAG is substantial and achievable without foundation-model training, but the method depends on costly expert curation and an unvalidated internal pipeline, so a small-scale pilot on one domain is warranted before production adoption.", "technique": "Structured experience-corpus engineering for agent RAG", "family": "recuperacao_e_rag", "recommendation": "testar", "published": "2026-09-29", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.37673", "independent_implementations": 2, "total_implementations": 3, "stars": 132, "citations": null, "claims": [ { "claim_id": "claim-01", "statement": "It replaces naive retrieval over raw work records with a curated experience corpus (rules, constraints, negative examples, corner cases, callable skills) built from practitioner records and interviews through a nine-layer extraction process. The cost is primarily human and engineering effort: interviews, semantic alignment, cross-review, and ongoing consolidation of 1,576 source files into tens of thousands of curated assets, rather than GPU or model-training spend. It can fail when expert time is unavailable, when domains drift so curated rules and boundaries go stale, or when the reported gains do not transfer since evaluation is internal, sample-based (20 practitioners), and lacks independent replication or public tooling.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.37673", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "On internal multi-domain tasks, the KUPAS MASTER agent scored 89.58 versus 79.75 for raw-corpus RAG and 70.63 for the base model under common scoring criteria.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.37673", "source_page": null, "source_excerpt": "" } ], "exposure_map": [ { "dimension": "quality", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compute", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "latency", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "operations", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compatibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "security", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "data_and_training", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "reproducibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" } ], "risks": [], "minimum_test": [], "open_questions": [], "independent_tests": [], "lab_reproductions": [], "report_available": false, "validation_tier": "unknown", "evidence_tier": "unknown", "infrastructure_basis": "unknown", "equations": [], "equations_status": "not_fetched", "core_kind": "", "equations_fetched_at": "", "equations_source_url": "", "equations_source_sha256": "", "exposures_reviewed_at": "", "exposures_source_sha256": "" }