randRuFlo 9b46f57c58 feat(metaharness): @metaharness/darwin as a regular dependency, used in-process (ADR-526); release 3.74.0 ...

- darwin moves to dependencies (ADR-526 amends ADR-150 rule 2 for darwin
  only); the smoke gate, doctor check and clean-install test accept it.
- src/metaharness/darwin-lib.ts: cached import; metaharness_bench verify and
  every metaharness_gepa op run in-process with the scripts' exact payloads
  and exit codes, falling back to the scripts when darwin does not import.
- ruflo-metaharness 0.5.0: bench.mjs verify imports darwin directly (bare
  import, no tool-time install) and falls back to the installed darwin CLI.
- Parity test vs the plugin scripts; benchmark script + measured runs.
- Version 3.74.0.

Co-Authored-By: RuFlo <ruv@ruv.net>

2026-11-08 22:01:25 -05:00

..

multi-agent

darwin: iter 11 advance — nfcorpus bm25 weight 0.7→0.4, ndcg10 0.3372→0.3443 (+0.71%)

2026-06-26 16:58:54 -04:00

runs

feat(metaharness): @metaharness/darwin as a regular dependency, used in-process (ADR-526); release 3.74.0

2026-11-08 22:01:25 -05:00

submission

docs(gaia): submission reproducibility package (stable 34/53 config) (#2213)

2026-05-28 21:02:02 -04:00

tdd-repair

feat(testgen): Test-Driven Repair via headless claude -p (ADR-175-inspired)

2026-06-22 17:15:32 -04:00

BEIR-MATRIX.md

feat(beir): 4-dataset BEIR (SciDocs joins) + config divergence finding (ADR-091, 3.10.30)

2026-05-31 03:17:53 -04:00

context-delegation.md

feat(core): governed context delegation with reproducible benchmark proof (#4648)

2026-11-01 14:13:37 -05:00

darwin-runs.jsonl

darwin: iter 23 log — meta memoryUsefulness 40→46, halt counter reset

2026-06-26 18:28:00 -04:00

evolution-log.jsonl

log: darwin iter 13 result

2026-06-26 17:17:24 -04:00

guidance-baseline.json

perf(guidance): Phase 1 — hot-path cleanup + scaling benchmark infrastructure

2026-05-21 23:21:29 -04:00

guidance-phase-1.json

perf(guidance): Phase 1 — hot-path cleanup + scaling benchmark infrastructure

2026-05-21 23:21:29 -04:00

guidance-quantization-m4.json

perf(guidance): M4 quantization — 1-bit RaBitQ-style signatures, 2.7x at N=1000

2026-05-21 23:51:48 -04:00

guidance-retriever-scale-baseline-filtered.json

perf(guidance): M3 substrate — packed embedding matrix + filtered-query bench coverage

2026-05-21 23:27:23 -04:00

guidance-retriever-scale-baseline.json

perf(guidance): Phase 1 — hot-path cleanup + scaling benchmark infrastructure

2026-05-21 23:21:29 -04:00

guidance-retriever-scale-m4.json

perf(guidance): M4 quantization — 1-bit RaBitQ-style signatures, 2.7x at N=1000

2026-05-21 23:51:48 -04:00

guidance-retriever-scale-phase-1.json

perf(guidance): Phase 1 — hot-path cleanup + scaling benchmark infrastructure

2026-05-21 23:21:29 -04:00

guidance-retriever-scale-phase-3-m3.json

perf(guidance): M3 substrate — packed embedding matrix + filtered-query bench coverage

2026-05-21 23:27:23 -04:00

guidance-retriever-scale-phase-3-m3v2-filtered.json

perf(guidance): M3 substrate — packed embedding matrix + filtered-query bench coverage

2026-05-21 23:27:23 -04:00

guidance-retriever-scale-phase-3-m3v2.json

perf(guidance): M3 substrate — packed embedding matrix + filtered-query bench coverage

2026-05-21 23:27:23 -04:00

managed-cycle-admission.json

feat: managed workflows, unified console runs and verified advisory learning (#4471)

2026-10-24 22:42:09 -04:00

managed-dynamic-workflows.json

feat: managed workflows, unified console runs and verified advisory learning (#4471)

2026-10-24 22:42:09 -04:00

managed-learning-console.json

feat: managed workflows, unified console runs and verified advisory learning (#4471)

2026-10-24 22:42:09 -04:00

rvagent-baseline.json

feat(rvagent): #ADR-129 — full rvagent integration (4 phases) (#2123)

2026-05-24 14:32:43 -04:00