Skip to content

armgit / ruflo

Public Imported from GitHub
0

🌊 The original agent harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, federation, vector RAG integration, and native Claude Code / Codex / Hermes and many more Integrated

2.2 KB
Raw Blame

Darwin capability evolution β€” plan

Branch: darwin/capability-evolution-2026-06-26 Started: 2026-06-26

Goal

Drive ruflo capabilities toward SOTA across the dimensions we already benchmark, using a /loop 5m autonomous loop. Each tick spawns one claude -p (headless, Read/Edit/Bash only, --max-budget-usd capped) to do a single optimization cycle, so this conversation stays focused on orchestration and the per-tick spend is bounded.

Per-tick contract

A single tick = one claude -p invocation that does end-to-end:

  1. Read docs/darwin/log.jsonl β€” last N entries, find current champion scores per dimension.
  2. Pick the worst-relative-to-SOTA dimension. SOTA baselines: BEIR NFCorpus β€” nDCG@10 β‰₯ 0.36 (state-of-the-art hybrid) BEIR ArguAna β€” nDCG@10 β‰₯ 0.55 BEIR SciFact β€” nDCG@10 β‰₯ 0.74 BEIR TREC-COVID β€” nDCG@10 β‰₯ 0.78 GAIA L1 β€” exact-match β‰₯ 0.62 (LangGraph reference) ADR coverage β€” adr-index storage success β‰₯ 0.99
  3. Propose ONE targeted change (parameter tune, prompt rewrite, dep bump, algorithm swap). Keep it small enough that a benchmark subset can score it in ≀4 minutes.
  4. Apply, run the relevant benchmark/audit: BEIR β†’ node scripts/bench-beir.mjs <dataset> --top-k 10 ADR β†’ node plugins/ruflo-adr/scripts/import.mjs --dry-run OIA β†’ npx ruflo metaharness oia-audit --format json
  5. Compare delta to prior champion for that dimension: Ξ” > 0 β†’ commit, update champion, log success Ξ” ≀ 0 β†’ revert, log noImprovement benchmark error β†’ revert, log error
  6. Append one JSONL line to docs/darwin/log.jsonl with: { iter, ts, dimension, change, deltaScore, action, commit }

Termination

  • 3 consecutive iterations without Ξ” > 0 across any dimension β†’ stop
  • Or explicit user stop

Spend cap per tick

claude -p --max-budget-usd 0.50 --model haiku for routine ticks. Escalate to sonnet only when haiku reports "task too complex" 3x in a row.

SOTA-proof

A dimension is "proven SOTA" when:

  • It exceeds the baseline above by β‰₯1%
  • The benchmark run is reproducible (3 consecutive runs within 1Οƒ)
  • The git commit is signed and witnessed