Starstruk Logo MarkStarstruk

v3.24.0

parruda/swarmv3.24.0Jul 5, 2026by ruvnet

AI Summary

Implements a self-learning flywheel that can improve retrieval policies over time using verified, signed champions and shadow-first promotion strategies.

Key Highlights

  • Verified retrieval improvement auto-applied via signed Ed25519 config champion.
  • Self-optimizing flywheel that compounds verified policy improvements.
  • Shadow-first promotion with drift canary and auto-rollback capabilities.
  • Meta-learning biases search toward historically high-payoff policy axes.

New Features

  • Signed config champion for baseline improvements
  • Self-optimizing flywheel (opt-in)
  • Shadow-first promotion and drift canary
  • Meta-learning optimizer

Full Release Notes

## ruflo 3.24.0 ā€” The Self-Learning Flywheel

ruflo can now improve one of its own operating policies over time and **prove** each improvement is real ā€” not marketing. Merges #2572 (ADR-176 self-optimizing flywheel + ADR-177 signed config propagation).

šŸ“– **Full write-up (plain language + technical + usage + upgrade notes):** https://gist.github.com/ruvnet/f8e2851fd307df5d5de7b5c70c37fa0c

### What's new
- **Verified retrieval improvement, auto-applied to every install.** A **signed** config champion (Ed25519 + RVFA) is adopted on startup, **fail-closed** on authenticity *and* suitability. Better retrieval defaults out of the box, +0.0738 nDCG@3 over the previously-tuned baseline. No re-init needed.
- **Self-optimizing flywheel (opt-in, $0 default).** The background daemon compounds *verified* retrieval-policy improvements: each generation reads the persisted champion as baseline, gates a candidate on a **frozen** held-out with a **significance** test (`accept/v1+sig`) + human-relevance guard + a **separate** canary, and on promotion advances the champion so the next tick builds on it. Winners accumulate into a signed, **independently-replayable** lineage back to an immutable root (git-for-operating-policies).
- **Shadow-first / no auto-serve + drift canary.** Promoted champions serve only after a one-generation shadow delay; a canary re-scores on the evolving store each tick and **auto-rolls-back** regressions.
- **Meta-learning.** The optimizer biases its search toward policy axes with measured historical payoff.
- **Proof, not assertion.** Receipt bundles replay independently *without trusting our logs*; a CI guard keeps the shipped evidence valid on every PR.

Demonstrated live: two real, significant, compounding promotions (self-retrieval RR **0.496 ā†’ 0.758 ā†’ 0.847**), human relevance preserved, zero human intervention.

> **Honest scope:** the flywheel's compounding gains are on a **self-supervised** retrieval benchmark, gated so human-labeled relevance does not regress ā€” not a claim that human relevance improved generation-over-generation (held flat by design). The auto-applied one-shot champion *was* tuned on human-labeled relevance.

### Upgrade
```bash
npx ruflo@latest        # or npx ruflo@3.24.0
```
Backwards-compatible / additive Ā· signing keys unchanged Ā· the flywheel is **off unless** `RUFLO_HARNESS_LOOP=1`.

**Packages:** `@claude-flow/cli@3.24.0` Ā· `claude-flow@3.24.0` Ā· `ruflo@3.24.0`

šŸ¤– Generated with [RuFlo](https://github.com/ruvnet/ruflo)