{"record": {"aliases": ["llms-benchmark"], "artifacts": [{"id": "gh:diovoib/llms-benchmark", "kind": "repo", "url": "https://github.com/diovoib/llms-benchmark"}], "categories": ["Agents", "General AI", "Language & Knowledge"], "description": {"en": "Handful of tasks for manual testing of local LLM tool calling and agent loop behavior."}, "documents": [{"document_type": "registry_page", "id": "claire_radar:https://github.com/diovoib/llms-benchmark", "source": "claire_radar", "source_url": "https://github.com/diovoib/llms-benchmark", "title": "Local agent benchmark (tools + loop + coding)"}], "evidence_summary": {"document_count": 1, "model_count": null, "model_count_basis": null}, "key": "claire-radar:github:diovoib/llms-benchmark", "modality": null, "name": "Local agent benchmark (tools + loop + coding)", "openness": {"code_license": null, "data_license": null, "evidence": [], "status": "unknown"}, "provenance": {"admission_policy_version": "2026-09-09.1", "confidence": "Medium", "crawl_bundle": "claire_radar_2026-10-14", "crawled_at": "2026-10-14T09:44:51Z", "data_status": "primary-source-candidate", "display_eligible": "false", "origin_record_id": "bm_llms-benchmark_268bf6b4", "origin_source": "github", "recognition_confidence": "0.4", "record_sha256": "36fd3e301d82c37161191f81654617b4bb47f76e8ee2e20a62233b9de45393a9", "relation": "unclear", "review_model": "deepseek-v4-pro", "review_state": "ai-name-audit-deferred", "reviewed_at": "2026-09-14T00:43:53.625753Z", "source_url": "https://github.com/diovoib/llms-benchmark"}, "publisher": null, "released": "2026-09-10", "released_reference": {"basis": "benchmark_release", "source_key": "claire-radar:github:diovoib/llms-benchmark", "source_url": "https://github.com/diovoib/llms-benchmark"}, "schema_version": 1, "sizes": [], "slug": "claire-radar-github-diovoib-llms-benchmark", "source": "claire_radar", "source_benchmark_id": "github:diovoib/llms-benchmark", "source_metadata": {"claire_radar": {"adoption30d": null, "aliases": ["llms-benchmark"], "annotation": "Unknown", "applicationDomains": ["General AI"], "area": "Language & Knowledge", "attention": {"asOf": "2026-09-11", "attemptedAt": "2026-09-11T06:43:09.447014Z", "githubRepo": "https://github.com/diovoib/llms-benchmark", "githubScope": "benchmark_repo", "githubStars": 0, "hfDailySubmittedAt": null, "hfDataset": null, "hfDatasetDownloads": null, "hfDatasetLikes": null, "hfPaperUpvotes": null, "hfPaperUrl": null, "observedAt": "2026-09-11T06:43:09.447014Z", "signalStatus": {"githubStars": {"observedAt": "2026-09-11T06:43:09.447014Z", "state": "fresh"}, "hfDatasetDownloads": {"state": "not_applicable"}, "hfDatasetLikes": {"state": "not_applicable"}, "hfPaperUpvotes": {"state": "not_applicable"}}, "status": "fresh"}, "attentionForecast": {"confidence": "Low", "horizon": "7d", "reason": "Not a public benchmark, so initial attention is expected to be negligible.", "score": 0}, "capabilities": [], "confidence": "Medium", "construction": "Unknown", "constructionDetail": "Unknown — the indexer does not infer construction details without explicit source evidence.", "copyGeneration": {"generatedAt": "2026-09-14T00:43:53.625753Z", "inputHash": "cf613b075b0e538d6ca5baf2ac619ae248a9ae6b9d19ac0a20e2814837408641", "model": "deepseek-v4-pro"}, "curation": {"admissionPolicyVersion": "2026-09-09.1", "canonicalNameEvidence": "# Local agent benchmark (tools + loop + coding)", "canonicalNameSource": "official_readme", "decisionReason": "Local test harness with manual judge interpretation; not a released benchmark with stable scoring contract.", "evaluationEvidence": {"reuse": "```text python run.py run python run.py run --verbose ```", "scoring": "The bench catches the obvious things — wrong tool, bad JSON, a template leaking into the chat — but it does not decide whether asking for a missing city was reasonable, or whether the code review was self-adoration.", "task": "This directory is for evaluating a **small local LLM** as an agent engine: the main emphasis here is whether it correctly calls the tools the harness exposes to it, whether it can organize a **tool** → **result** → **next-step** loop, and whether on a longer coding task it can check its own work, make actual progress, and decide when the task is done."}, "model": "deepseek-v4-pro", "reviewedAt": "2026-09-14T00:43:53.625753Z", "state": "ai-name-audit-deferred"}, "dataStatus": "primary-source-candidate", "demo": false, "description": "Handful of tasks for manual testing of local LLM tool calling and agent loop behavior.", "discoverySources": [{"id": "github:diovoib/llms-benchmark", "type": "github", "url": "https://github.com/diovoib/llms-benchmark"}], "displayEligible": false, "domainCuration": {"method": "rules-v1", "state": "auto"}, "evaluationMode": "score_submission", "evidence": {"reasonCodes": ["discovered via github", "benchmark term in abstract", "evaluation protocol evidence", "public artifact URL", "no explicit benchmark release evidence"], "snippet": "llms-benchmark The bench scores small local LLM models as an agent over an OpenAI-compatible chat/completions API: whether it calls harness tools correctly, can run a tool-result-next-step loop, and makes progress on a longer coding task."}, "familyId": "bmf_018f7dc40595", "firstSeenAt": "2026-09-11", "heat": null, "id": "bm_llms-benchmark_268bf6b4", "indexedAt": "2026-09-11T06:20:17.513010Z", "industrySectors": [], "links": {"code": "https://github.com/diovoib/llms-benchmark", "project": "http://127.0.0.1:8080/v1`"}, "metrics": [], "motivation": "llms-benchmark The bench scores small local LLM models as an agent over an OpenAI-compatible chat/completions API: whether it calls harness tools correctly, can run a tool-result-next-step loop, and makes progress on a longer coding task.", "name": "Local agent benchmark (tools + loop + coding)", "oneLine": "Handful of tasks for manual testing of local LLM tool calling and agent loop behavior.", "paperTitle": "llms-benchmark", "primaryDomain": "General AI", "publication": {"evidence": null, "evidenceLevel": "unverified", "evidenceUrl": "https://github.com/diovoib/llms-benchmark", "source": "arxiv-metadata", "status": "unverified", "venue": null, "verifiedAt": "2026-09-11T06:20:17.513010Z"}, "publications": [], "ranking": {"30d": {"confidence": "Low", "coverage": 0.55, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 30}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.0, "value": 0}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 138, "score": 23}, "method": "attention-ranking-v13", "rank": 138, "score": 23}, "90d": {"confidence": "Low", "coverage": 0.55, "growth": {"components": {"githubStars": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfDatasetDownloads": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}, "hfPaperUpvotes": {"normalization": "percentile", "normalized": null, "percentile": null, "value": null}}, "confidence": "Low", "coverage": 0.0, "elapsedDays": null, "historyDate": null, "rank": null, "score": null, "windowDays": 90}, "level": {"components": {"githubStars": {"normalization": "log-max", "normalized": 0.0, "value": 0}, "hfDatasetDownloads": {"normalization": "log-max", "normalized": null, "value": null}, "hfPaperUpvotes": {"normalization": "log-max", "normalized": null, "value": null}}, "confidence": "Low", "coverage": 0.55, "rank": 338, "score": 23}, "method": "attention-ranking-v13", "rank": 338, "score": 23}}, "readiness": "Runnable", "recognitionConfidence": 0.4, "relation": "unclear", "releasedAt": "2026-09-10", "source": {"authors": ["diovoib"], "categories": [], "id": "github:diovoib/llms-benchmark", "publicSignals": {"githubStars": 0}, "title": "llms-benchmark", "type": "github", "url": "https://github.com/diovoib/llms-benchmark"}, "sourceUpdatedAt": "2026-09-10T19:49:32Z", "topics": ["Agents"], "venueAttempts": [], "whyItMatters": "Not a formal benchmark with comparable scoring; lacks standardized protocol."}}}, "schema_version": 1, "scores_by_source": {}, "siblings": []}
