{
  "schema_version": 1,
  "builder_plan": {
    "scenario": "A workflow needs specialist review, but extra agents may add more cost than useful work.",
    "change": "Test the proposed division of work against a single-agent implementation.",
    "baseline": "One agent with the same tools and an equal total token and time 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": "Accepted outputs, coordination failures, wall-clock time, and total cost across all agents.",
    "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": "2610.21791",
  "title": "RSIFlow: Adaptive Self-Improvement Workflows with an Evolving Meta-Agent",
  "summary": "RSIFlow replaces fixed, hand-designed self-improvement loops with a meta-agent that diagnoses bottlenecks from trajectories and dynamically selects component-specific improvement workflows, distilling outcomes into reusable skills. It costs substantial additional LLM calls for diagnosis, workflow selection, and meta-agent harness evolution, increasing per-task token spend, latency, and orchestration complexity relative to single-pass agent runs. After adoption it can fail through compounding errors if the meta-agent distills spurious lessons from noisy feedback, through evaluation-benchmark overfitting, and through unbounded recursive refinement that inflates API cost without proportional quality gain.",
  "rationale": "The framework is implementable against third-party APIs without training infrastructure, but its recursive improvement loops carry high and workload-dependent token costs and a real risk of self-reinforcing errors, so it warrants a bounded pilot rather than production adoption.",
  "technique": "Recursive self-improvement with evolving meta-agent orchestration",
  "family": "orquestracao_multiagente",
  "recommendation": "testar",
  "published": "2026-10-15",
  "as_of": "2026-10-18",
  "editorial_status": "indexed",
  "source_url": "https://arxiv.org/abs/2610.21791",
  "independent_implementations": 1,
  "total_implementations": 2,
  "stars": 132,
  "citations": null,
  "claims": [
    {
      "claim_id": "claim-01",
      "statement": "RSIFlow replaces fixed, hand-designed self-improvement loops with a meta-agent that diagnoses bottlenecks from trajectories and dynamically selects component-specific improvement workflows, distilling outcomes into reusable skills. It costs substantial additional LLM calls for diagnosis, workflow selection, and meta-agent harness evolution, increasing per-task token spend, latency, and orchestration complexity relative to single-pass agent runs. After adoption it can fail through compounding errors if the meta-agent distills spurious lessons from noisy feedback, through evaluation-benchmark overfitting, and through unbounded recursive refinement that inflates API cost without proportional quality gain.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.21791",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "claim_id": "claim-02",
      "statement": "Reports state-of-the-art average performance across six benchmarks and two backbone models, with ablations attributing gains to adaptive workflow orchestration and meta-harness evolution; no multiplicative factor is stated.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2610.21791",
      "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": ""
}