{ "schema_version": 1, "builder_plan": { "scenario": "An assistant chooses the wrong API or sends arguments your product cannot use.", "change": "Apply the proposed tool-selection or argument-generation change in a sandbox.", "baseline": "Your current agent with the same tool schemas and task budget.", "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": "Correct completed actions, invalid calls, retries, latency, and cost per completed task.", "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": "2610.21267", "title": "TRACER: Trajectory-Aware Tool Retrieval for LLM Agents", "summary": "It replaces exposing the full tool registry in the prompt with a retrieval front-end that indexes observed tool-call sequences in a trie and ranks a small candidate subset using a learned blend of empirical and contextual priors. It costs building and maintaining the trajectory index and the blending model, plus a retrieval step per turn, and its benefit depends on having logged tool-call sequences representative of production usage. It can fail when production trajectories diverge from indexed ones (cold-start or novel workflows), where pruning the registry to five tools may omit the correct tool, and the paper notes three model-environment pairs where accuracy trails full-registry selection by up to 2.2 points.", "rationale": "The token savings and accuracy gains are directly relevant to a small team paying per-token API costs, but the method requires logged tool-call trajectories and its advantage over simpler embedding-based tool retrieval is unverified on the reader's workload, so a scoped pilot is warranted before adoption.", "technique": "Trajectory-aware trie retrieval of tool subsets", "family": "uso_de_ferramenta", "recommendation": "testar", "published": "2026-10-15", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2610.21267", "independent_implementations": 0, "total_implementations": 1, "stars": 10, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "It replaces exposing the full tool registry in the prompt with a retrieval front-end that indexes observed tool-call sequences in a trie and ranks a small candidate subset using a learned blend of empirical and contextual priors. It costs building and maintaining the trajectory index and the blending model, plus a retrieval step per turn, and its benefit depends on having logged tool-call sequences representative of production usage. It can fail when production trajectories diverge from indexed ones (cold-start or novel workflows), where pruning the registry to five tools may omit the correct tool, and the paper notes three model-environment pairs where accuracy trails full-registry selection by up to 2.2 points.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.21267", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "On a mixed retail-and-telecom environment, TRACER exposes 5 of 128 tools, cutting input tokens per step by 80% while raising GPT-5.4 tool-selection accuracy by 4.7 points; across six environments and four models it improves selection accuracy in 21 of 24 pairs, with average gains of 6 to 10 points.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2610.21267", "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": "" }