ruvnet/metaharness
50.3
Adequate · 17 November 2026
145.5k
lines of production code
TypeScript
with JavaScript, Rust
1
measurement over time
What this system is
Behavioural changes
1 commit (0 fixes) modifying crates/kernel-napi
A change to existing behaviour in crates/kernel-napi — 1 commit, 2 files: “feat(iter-1): scaffold Cargo + npm workspaces, kernel stubs, CI/publish/security pipeline”.
crates/kernel-napi · medium confidence · unverified
1 commit (0 fixes) modifying kimi-k3-harness, kimi-k3-harness/crates/k3-kernel-bench, kimi-k3-harness/crates/k3rs
A change to existing behaviour in kimi-k3-harness, kimi-k3-harness/crates/k3-kernel-bench, kimi-k3-harness/crates/k3rs — 1 commit, 48 files: “feat: kimi-k3 harness, C-parity Rust/WASM K3 engine, AgentRadio + NOOA clones, darwin/flywheel optimizations — all operational, replay-verified (\#170)” (refs \#170).
kimi-k3-harness, kimi-k3-harness/crates/k3-kernel-bench, kimi-k3-harness/crates/k3rs · medium confidence · unverified
1 commit (0 fixes) modifying packages/darwin-mode/bench/ablation
A change to existing behaviour in packages/darwin-mode/bench/ablation — 1 commit, 1 file: “bench(darwin): Poincaré-vs-Euclidean niche ablation — honest result (ADR-095, Gap 1)”.
packages/darwin-mode/bench/ablation · medium confidence · unverified
1 commit (0 fixes) modifying packages/darwin-mode/bench/confirmation
A change to existing behaviour in packages/darwin-mode/bench/confirmation — 1 commit, 10 files: “feat(darwin): independent contrastive confirmation with public pilot evidence”.
packages/darwin-mode/bench/confirmation · medium confidence · unverified
1 commit (0 fixes) modifying packages/darwin-mode/bench/translate
A change to existing behaviour in packages/darwin-mode/bench/translate — 1 commit, 2 files: “feat(translate): chunked-iterative-synthesis harness (ADR-191) — bounded greenfield Python→Rust translation driver with tiered routing + per-chunk Rust-test gates; sample TimesFM 4-phase manifest; offline --plan mode”.
packages/darwin-mode/bench/translate · medium confidence · unverified
1 commit (0 fixes) modifying packages/darwin-mode/src/gepa
A change to existing behaviour in packages/darwin-mode/src/gepa — 1 commit, 5 files: “feat(darwin): 0.8.0 — ship the GEPA engine as @metaharness/darwin/gepa (pluggable evaluator) + cand-6 genome (\#74)” (refs \#74).
packages/darwin-mode/src/gepa · medium confidence · unverified
1 commit (0 fixes) modifying packages/evals-extract
A change to existing behaviour in packages/evals-extract — 1 commit, 14 files: “feat(evals-extract): @metaharness/evals-extract — structured-extraction flywheel adapter (\#98)” (refs \#98).
packages/evals-extract · medium confidence · unverified
1 commit (0 fixes) modifying packages/evals-servedmodel
A change to existing behaviour in packages/evals-servedmodel — 1 commit, 12 files: “feat(ruvllm): @metaharness/evals-servedmodel — served-model flywheel adapter (ADR-234, closes \#109 SYNTHETIC-proven) (\#110)” (refs \#109, \#110).
packages/evals-servedmodel · medium confidence · unverified
1 commit (0 fixes) modifying packages/evals-sql
A change to existing behaviour in packages/evals-sql — 1 commit, 14 files: “feat(evals-sql): @metaharness/evals-sql — flywheel domain adapter (multi-vertical scale-up) (\#97)” (refs \#97).
packages/evals-sql · medium confidence · unverified
1 commit (0 fixes) modifying services/apicompletions, services/apicompletions/functions/aggregateUsage
A change to existing behaviour in services/apicompletions, services/apicompletions/functions/aggregateUsage — 1 commit, 35 files: “feat: ADR-228 GEPA Distilled Executor Genome — offline evolution of the cheap-executor policy (\#64)” (refs \#64).
services/apicompletions, services/apicompletions/functions/aggregateUsage · medium confidence · unverified
10 commits (3 fixes) modifying examples-packages
A change to existing behaviour in examples-packages — 10 commits (3 fixs), 131 files: “docs: P0-5 — mark example-README transcripts illustrative + fix harness cmd (issue \#4)”; “feat(examples): bespoke per-SDK agent bundles — scaffolds emit agents + MCP policy (0.2.0) (\#29)”; “feat(examples): iter 148 — 19th example package @metaharness/github-actions (live) + gist + docs”; and 7 more (refs \#4, \#29, \#482, \#490, \#19).
examples-packages · medium confidence · unverified
11 commits (1 fix) modifying apps/web-ui/src
A change to existing behaviour in apps/web-ui/src — 11 commits (1 fix), 3 files: “refactor(brand): iter 118 — mintagent → openharness rename + "Open Harness Studio" UI + Studio SEO”; “feat(web-ui): searchable categorized Compose picker + first-class Darwin mode (ADR-171)”; “feat(web-ui): social link-preview card + surface score/router in hero”; and 8 more (refs \#58).
apps/web-ui/src · medium confidence · unverified
11 commits (5 fixes) modifying packages/host-hermes
A change to existing behaviour in packages/host-hermes — 11 commits (5 fixs), 5 files: “feat(iter-8): harness sign/verify/doctor CLI + MCP tool registry + 6 package READMEs + GCP setup script”; “dream(host-adapters): fix hermes YAML mapping-key injection, all 3 codegen paths (evaluated) (\#188)”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; and 8 more (refs \#188, \#439, \#482, \#7, \#490, …).
packages/host-hermes · medium confidence · unverified
12 commits (2 fixes) modifying packages/projects/bench
A change to existing behaviour in packages/projects/bench — 12 commits (2 fixs), 28 files: “projects: scaled budget-capped multi-seed frontier bake-off bench”; “bench+opt(projects): hard-task proposer bake-off — frontier earns its cost; severity fix”; “feat: GLM open-frontier lane + zero-day harness optimization (39x-\>1 semgrep)”; and 9 more.
packages/projects/bench · medium confidence · unverified
120 commits (26 fixes) modifying packages/create-agent-harness/src
A change to existing behaviour in packages/create-agent-harness/src — 120 commits (26 fixs), 47 files: “feat(metaharness): add @metaharness/darwin as a dependency + bump scaffold version to ^0.2.2”; “feat(cli): iter 135 — \harness plugin-init\ backfills .claude-plugin/plugin.json for legacy harnesses”; “feat(proxy): route launched clients with worktree policies”; and 117 more (refs \#145, \#22, \#121, \#4, \#124, …).
packages/create-agent-harness/src · medium confidence · unverified
13 commits (3 fixes) modifying apps/web-ui/public
A change to existing behaviour in apps/web-ui/public — 13 commits (3 fixs), 2 files: “feat(leaderboard): rank vs REAL SWE-bench Lite entries + cost estimated from disclosed models; load from assets/JSON”; “leaderboard: remove SWE-bench Pro tab/data from cost-pareto page (Pro is a cheap-harness dead-end per §45 — turn-budget cliff)”; “ui(welcome): first-visit welcome splash for the leaderboard — brief Pareto intro + animated frontier (line draws in, frontier dots pulse, dominated dots fade); Explore/How-it-works buttons, localStorage-dismissed, ?welcome=1 to replay”; and 10 more.
apps/web-ui/public · medium confidence · unverified
13 commits (3 fixes) modifying packages/projects/src
A change to existing behaviour in packages/projects/src — 13 commits (3 fixs), 23 files: “feat: Tier 1-2 — pipeline+CLI, taint oracle, multi-lang verifier, deeper learning loop, CI gate, CodeQL shell”; “bench+opt(projects): hard-task proposer bake-off — frontier earns its cost; severity fix”; “feat: GLM open-frontier lane + zero-day harness optimization (39x-\>1 semgrep)”; and 10 more (refs \#169).
packages/projects/src · medium confidence · unverified
13 commits (5 fixes) modifying packages/redblue
A change to existing behaviour in packages/redblue — 13 commits (5 fixs), 37 files: “fix(redblue): close indirect\_prompt\_injection structured-field bypass (\#186)”; “fix(redblue): isolate H1 taxonomy cache in tests (REDBLUE\_H1\_CACHE override)”; “feat(redblue): umbrella CLI integration + SEO/intro + 0.1.1 (mirror weight-eft)”; and 10 more (refs \#186, \#181, \#190, \#482, \#182).
packages/redblue · medium confidence · unverified
148 commits (17 fixes) modifying packages/darwin-mode/bench/swebench
A change to existing behaviour in packages/darwin-mode/bench/swebench — 148 commits (17 fixs), 304 files: “docs(darwin/swebench): ADR-143 — closed-loop vs open-loop A/B (16% vs 12%, within noise at n=25)”; “feat(darwin/ADR-149): closed-loop repair DOUBLES SWE-bench Lite resolve-rate — 46/300 = 15.3% \[11.7,19.8\]”; “feat(darwin/swebench): system-message format contract + --slice (unblocks small-context local models, ADR-150)”; and 142 more (refs \#113, \#105, \#100, \#94, \#101, …).
packages/darwin-mode/bench/swebench · medium confidence · unverified
15 commits (1 fix) modifying .codex
A change to existing behaviour in .codex — 15 commits (1 fix), 30 files: “docs: iter 91 — propagate diag --bundle across user-facing surfaces”; “feat(examples): iter 148 — 19th example package @metaharness/github-actions (live) + gist + docs”; “fix: integrate Dream changes with security, transport and upgrade regressions covered”; and 12 more (refs \#169).
.codex · medium confidence · unverified
15 commits (1 fix) modifying packages/darwin-mode/src/security
A change to existing behaviour in packages/darwin-mode/src/security — 15 commits (1 fix), 23 files: “feat(darwin): implement Darwin Shield defensive zero-day harness (ADR-155)”; “feat: Tier 1-2 — pipeline+CLI, taint oracle, multi-lang verifier, deeper learning loop, CI gate, CodeQL shell”; “feat(darwin): real Semgrep DRIVES the evolutionary loop (ADR-155 Addendum A Phase 2 capstone)”; and 12 more.
packages/darwin-mode/src/security · medium confidence · unverified
15 commits (2 fixes) modifying packages/kernel-js/src
A change to existing behaviour in packages/kernel-js/src — 15 commits (2 fixs), 9 files: “feat(iter-1): scaffold Cargo + npm workspaces, kernel stubs, CI/publish/security pipeline”; “dream(host-adapters): \#385 close SessionLog.append()/fork() TOCTOU race”; “fix(kernel,generator): generated harness is now a runnable/testable npx package”; and 12 more (refs \#385, \#20, \#21, \#22, \#27, …).
packages/kernel-js/src · medium confidence · unverified
152 commits (48 fixes) modifying scripts
A change to existing behaviour in scripts — 152 commits (48 fixs), 52 files: “test(proxy-pin-drift): make the PEM assertion EOL-agnostic — main is red on Windows”; “feat(iter-1): scaffold Cargo + npm workspaces, kernel stubs, CI/publish/security pipeline”; “feat(gcp): workers self-report results to Firestore + cluster supervise loop”; and 144 more (refs \#96, \#213, \#472, \#427, \#84, …).
scripts · medium confidence · unverified
17 commits (3 fixes) modifying apps/web-ui/src/generator
A change to existing behaviour in apps/web-ui/src/generator — 17 commits (3 fixs), 18 files: “feat(web-ui): add Sonnet 4.6 + Opus 4.8, make them default Scholar/Sage (ADR-171)”; “feat(web-ui): Darwin self-evolution + model-tier config in the Studio (ADR-171)”; “feat(web-ui): searchable categorized Compose picker + first-class Darwin mode (ADR-171)”; and 14 more (refs \#188, \#211, \#212, \#439, \#169, …).
apps/web-ui/src/generator · medium confidence · unverified
18 commits (0 fixes) modifying apps/web-ui/src/components
A change to existing behaviour in apps/web-ui/src/components — 18 commits, 11 files: “feat(web-ui): add Sonnet 4.6 + Opus 4.8, make them default Scholar/Sage (ADR-171)”; “refactor(brand): iter 118 — mintagent → openharness rename + "Open Harness Studio" UI + Studio SEO”; “feat(web-ui): Darwin self-evolution + model-tier config in the Studio (ADR-171)”; and 15 more (refs \#443, \#169).
apps/web-ui/src/components · medium confidence · unverified
19 commits (5 fixes) modifying packages/darwin-mode/dist
A change to existing behaviour in packages/darwin-mode/dist — 19 commits (5 fixs), 268 files: “feat(darwin-mode): add numeric genome kind (ADR-272) (\#260)”; “feat: @metaharness/avo — governed autonomous variation runtime (ADR-251) + live provider validation (\#213)”; “feat(darwin/ADR-153): agentic loop at scale = 94/300 = 31.3% \[26.3,36.8\] on v4-pro — competitive + cheaper”; and 16 more (refs \#260, \#213, \#465, \#170, \#344, …).
packages/darwin-mode/dist · medium confidence · unverified
2 commits (0 fixes) modifying crates/kernel-wasm
A change to existing behaviour in crates/kernel-wasm — 2 commits, 1 file: “feat(iter-1): scaffold Cargo + npm workspaces, kernel stubs, CI/publish/security pipeline”; “Prime Agent integration: ADR-241/242 implemented — 10th host, kernel session+autonomous primitives (Rust+TS/wasm), RefineMutator (\#169)” (refs \#169).
crates/kernel-wasm · medium confidence · unverified
2 commits (0 fixes) modifying packages/autogenous
A change to existing behaviour in packages/autogenous — 2 commits, 10 files: “feat: add Autogenous MetaHarness adapter (\#208)”; “feat(autogenous): depend on the real autogenous and radio-moe packages (ADR-323) (\#497)” (refs \#208, \#497).
packages/autogenous · medium confidence · unverified
2 commits (0 fixes) modifying packages/darwin-mode/bench/polyglot
A change to existing behaviour in packages/darwin-mode/bench/polyglot — 2 commits, 3 files: “feat(darwin): polyglot execution-scored benchmark + route mutator to gemini-2.5-flash”; “bench(darwin): extend polyglot benchmark to 15 models (Chinese + Mistral)”.
packages/darwin-mode/bench/polyglot · medium confidence · unverified
2 commits (0 fixes) modifying packages/darwin-mode/bench/swebench/runs
A change to existing behaviour in packages/darwin-mode/bench/swebench/runs — 2 commits, 12 files: “feat(darwin/ADR-196): execution-trace localization A/B on HARD-25 — cracks 1 Opus give-up (+1), after fixing a silent trace-truncation bug (LEARNINGS §55)”; “feat(darwin/ADR-196 §62): trace-localize conformant A/B RAN — fire 31/35, K=3, +1.0pp within CI”.
packages/darwin-mode/bench/swebench/runs · medium confidence · unverified
2 commits (0 fixes) modifying packages/evals-hle
A change to existing behaviour in packages/evals-hle — 2 commits, 14 files: “docs(adr): ADR-236 — D1 SWE-bench domain run is an HONEST NULL + wire ADR-235 gate re-execution (\#100)”; “feat(evals-hle): @metaharness/evals-hle — HLE as a policy-evolution flywheel adapter (ADR-233) (\#89)” (refs \#100, \#89).
packages/evals-hle · medium confidence · unverified
2 commits (0 fixes) modifying packages/evals-toolcall
A change to existing behaviour in packages/evals-toolcall — 2 commits, 16 files: “feat(evals-toolcall): @metaharness/evals-toolcall — flywheel domain adapter (multi-vertical scale-up) (\#96)”; “Prime Agent integration: ADR-241/242 implemented — 10th host, kernel session+autonomous primitives (Rust+TS/wasm), RefineMutator (\#169)” (refs \#96, \#169).
packages/evals-toolcall · medium confidence · unverified
2 commits (0 fixes) modifying services
A change to existing behaviour in services — 2 commits, 20 files: “feat: ADR-228 GEPA Distilled Executor Genome — offline evolution of the cheap-executor policy (\#64)”; “chore: mark services/apicompletions as a DEPRECATED stale mirror → cognitum-one/meta-llm (\#66)” (refs \#64, \#66).
services · medium confidence · unverified
2 commits (1 fix) modifying packages/arc-agi-3-bench
A change to existing behaviour in packages/arc-agi-3-bench — 2 commits (1 fix), 23 files: “feat(arc): add governed AVO harness, ChatGPT adapter, and benchmarks (\#220)”; “test(arc-bench): fix the SECOND racy .partial poll — my earlier fix was incomplete” (refs \#220).
packages/arc-agi-3-bench · medium confidence · unverified
2 commits (1 fix) modifying packages/darwin-mode/bench/contrastive
A change to existing behaviour in packages/darwin-mode/bench/contrastive — 2 commits (1 fix), 9 files: “feat(darwin): bounded contrastive development API v0.11.0”; “fix(darwin): enforce total contrastive deadline and benchmark integration”.
packages/darwin-mode/bench/contrastive · medium confidence · unverified
2 commits (1 fix) modifying packages/darwin-mode/bench/selection
A change to existing behaviour in packages/darwin-mode/bench/selection — 2 commits (1 fix), 2 files: “fix: close ADR-index drift, rebuild stale darwin-mode dist, correct README/package.json claims (\#141)”; “feat(darwin): 0.14.0, optional ruv-fann surrogate pre-screen for numeric genomes (\#492)” (refs \#141, \#492).
packages/darwin-mode/bench/selection · medium confidence · unverified
2 commits (1 fix) modifying packages/darwin-mode/dist/gepa
A change to existing behaviour in packages/darwin-mode/dist/gepa — 2 commits (1 fix), 20 files: “fix: close ADR-index drift, rebuild stale darwin-mode dist, correct README/package.json claims (\#141)”; “feat(darwin): 0.8.0 — ship the GEPA engine as @metaharness/darwin/gepa (pluggable evaluator) + cand-6 genome (\#74)” (refs \#141, \#74).
packages/darwin-mode/dist/gepa · medium confidence · unverified
2 commits (1 fix) modifying packages/field-memory
A change to existing behaviour in packages/field-memory — 2 commits (1 fix), 12 files: “feat(field-memory): add governed packed attractor routing (\#202)”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)” (refs \#202, \#482).
packages/field-memory · medium confidence · unverified
2 commits (1 fix) modifying packages/oo-agents, packages/radio
A change to existing behaviour in packages/oo-agents, packages/radio — 2 commits (1 fix), 26 files: “feat: kimi-k3 harness, C-parity Rust/WASM K3 engine, AgentRadio + NOOA clones, darwin/flywheel optimizations — all operational, replay-verified (\#170)”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)” (refs \#170, \#482).
packages/oo-agents, packages/radio · medium confidence · unverified
2 commits (1 fix) modifying packages/turn-credit
A change to existing behaviour in packages/turn-credit — 2 commits (1 fix), 11 files: “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “feat: recursive turn credit (AgentOPSD, ADR-248) + four provable enhancements (ADR-249, calibration, credit-feedback) (\#176)” (refs \#482, \#176).
packages/turn-credit · medium confidence · unverified
22 commits (9 fixes) modifying packages/create-agent-harness/templates
A change to existing behaviour in packages/create-agent-harness/templates — 22 commits (9 fixs), 337 files: “feat(scaffold): iter 132 — propagate .claude-plugin/plugin.json to ALL 20 templates (was minimal-only)”; “feat(template): iter 80 (MILESTONE) — vertical:education, 17th vertical”; “feat(cli): multi-host scaffold + gate Claude files + bump vitest (GH \#10/\#11/\#12)”; and 19 more (refs \#10, \#23, \#24, \#26, \#161, …).
packages/create-agent-harness/templates · medium confidence · unverified
23 commits (5 fixes) modifying packages/flywheel
A change to existing behaviour in packages/flywheel — 23 commits (5 fixs), 37 files: “feat(flywheel): analyze reports cost-per-win economics (cost-per-accepted-task) (0.1.6) (\#107)”; “feat(flywheel): @metaharness/flywheel@0.1.1 + metaharness flywheel CLI (run\|replay\|graph) (\#84)”; “dream(flywheel-promotion): \#287 wire withSequentialEvidence into run.ts + replay.ts (evaluated) (\#288)”; and 20 more (refs \#107, \#84, \#287, \#288, \#306, …).
packages/flywheel · medium confidence · unverified
26 commits (6 fixes) modifying packages/bench/src/draco
A change to existing behaviour in packages/bench/src/draco — 26 commits (6 fixs), 14 files: “feat(draco): progress sidecar for live runs (.progress JSON)”; “fix(draco): per-request timeout + progress heartbeat for long live runs”; “feat(draco): self-consistency best-of-N selection arm (ADR-038 arm 2)”; and 23 more.
packages/bench/src/draco · medium confidence · unverified
27 commits (0 fixes) modifying packages/darwin-mode/bench
A change to existing behaviour in packages/darwin-mode/bench — 27 commits, 4 files: “feat(darwin): git-apply patch primitive + whole-file\>raw-diff finding — ADR-098 step-3 prep (ADR-124)”; “feat(darwin): micro-evolve noise floor + model×patchMode epistasis — honest stop (ADR-137)”; “feat(darwin): SWE runner generalizes to an external package (kernel-js) — ADR-131”; and 24 more.
packages/darwin-mode/bench · medium confidence · unverified
3 commits (0 fixes) modifying experiments
A change to existing behaviour in experiments — 3 commits, 24 files: “dream(flywheel-promotion): \#306 close meetsPromotionRule's noopRate ceiling lockout (evaluated) (\#307)”; “feat: kimi-k3 harness, C-parity Rust/WASM K3 engine, AgentRadio + NOOA clones, darwin/flywheel optimizations — all operational, replay-verified (\#170)”; “feat: recursive turn credit (AgentOPSD, ADR-248) + four provable enhancements (ADR-249, calibration, credit-feedback) (\#176)” (refs \#306, \#307, \#170, \#176).
experiments · medium confidence · unverified
3 commits (0 fixes) modifying packages/arc-agi-3
A change to existing behaviour in packages/arc-agi-3 — 3 commits, 25 files: “feat(arc): add governed AVO harness, ChatGPT adapter, and benchmarks (\#220)”; “feat: add ChatGPT ARC AGI 3 metaharness (\#218)”; “metaharness 0.4.21: ruvllm ^4.1.0, audit overrides (\#466), per-package provenance release (\#467)” (refs \#220, \#218, \#466, \#467).
packages/arc-agi-3 · medium confidence · unverified
3 commits (0 fixes) modifying packages/avo/bench
A change to existing behaviour in packages/avo/bench — 3 commits, 8 files: “feat: @metaharness/avo — governed autonomous variation runtime (ADR-251) + live provider validation (\#213)”; “bench(avo): ADR-251 preregistered 100-task SWE-bench gate — executed, ship=false on cost (\#217)”; “dream(flywheel-promotion): consolidate \#205/\#231/\#255 + close a 4th replay-trust gap (evaluated) (\#258)” (refs \#213, \#217, \#205, \#258).
packages/avo/bench · medium confidence · unverified
3 commits (0 fixes) modifying packages/darwin-mode/bench/cognition
A change to existing behaviour in packages/darwin-mode/bench/cognition — 3 commits, 11 files: “release(kernel): 0.1.4 — widen @ruvector/rvf peer; declare agenticow / tiny-dancer (\#473)”; “feat(cognition): memory-as-cognition A/B/C — agenticow branchable memory + Darwin evolve is NULL on a cheap model” (refs \#473).
packages/darwin-mode/bench/cognition · medium confidence · unverified
3 commits (0 fixes) modifying packages/darwin-mode/bench/livecodebench
A change to existing behaviour in packages/darwin-mode/bench/livecodebench — 3 commits, 34 files: “docs(bench): agentic-board integration adapters + roadmap (workflow w9811j59k, 6 boards)”; “feat(darwin): LiveCodeBench validated n=25 — single-shot deepseek = 16/25 (64%), TDR repair no lift (ADR-LCB)”; “feat(lcb): robust extractor + cost-cascade to reasoner at honest n=100 (ADR-LCB)”.
packages/darwin-mode/bench/livecodebench · medium confidence · unverified
3 commits (0 fixes) modifying packages/darwin-mode/bench/retort
A change to existing behaviour in packages/darwin-mode/bench/retort — 3 commits, 1 file: “feat(darwin-bench): retort metaharness greenfield solver”; “feat(retort): iteration-2 — fix timeout lever (multi-action + 20-min cap) + agenticow memory factor”; “feat(retort): iteration-3 — task-difficulty-aware routing; "no free lunch" confirmed (no beyond-SOTA)”.
packages/darwin-mode/bench/retort · medium confidence · unverified
3 commits (0 fixes) modifying packages/darwin-mode/bench/swebench/gepa
A change to existing behaviour in packages/darwin-mode/bench/swebench/gepa — 3 commits, 64 files: “feat: ADR-228 GEPA Distilled Executor Genome — offline evolution of the cheap-executor policy (\#64)”; “feat(gepa): Agenticow branch-memory in the GEPA candidate lifecycle (ADR-230 pilot) — reconciled onto \#64 (\#65)”; “feat(gepa): concurrent + gateway-backed overnight training loop; land cand-6 promotion (\#70)” (refs \#64, \#65, \#70).
packages/darwin-mode/bench/swebench/gepa · medium confidence · unverified
3 commits (0 fixes) modifying submissions/swe-bench-verified
A change to existing behaviour in submissions/swe-bench-verified — 3 commits, 5 files: “bench(§47): preserve Verified-500 cascade preds (499 records) — 55.6% gold-eval artifact (results.json to follow when eval exits)”; “bench(§47): capture Verified-500 official results.json (resolved\_ids) — 278/500=55.6% gold-eval artifact”; “docs(verified): assemble SWE-bench Verified submission package (REPORT + nested metadata + README/checklist) — 55.6% (278/500), cost labeled estimated; READY pending framing sign-off (no PR opened)”.
submissions/swe-bench-verified · medium confidence · unverified
3 commits (1 fix) modifying crates/kernel/benches
A change to existing behaviour in crates/kernel/benches — 3 commits (1 fix), 3 files: “feat(iter-2): 4 host adapters + minimal template + claude marketplace plugin + ts tests + rust benches”; “feat(iter-3): real Ed25519 witness signing, Codex skills, GCP secrets doc, template engine, security policies”; “fix(ci): bump rust toolchain to 1.88.0 + reformat + relax missing\_docs”.
crates/kernel/benches · medium confidence · unverified
3 commits (1 fix) modifying packages/aws-finops
A change to existing behaviour in packages/aws-finops — 3 commits (1 fix), 29 files: “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “ADR-168 + bootstrap @metaharness/aws-finops: Darwin primitives → AWS cost”; “aws-finops: real-tool oracle bench — terraform + checkov discrimination” (refs \#482).
packages/aws-finops · medium confidence · unverified
3 commits (1 fix) modifying packages/darwin-mode/dist/contrastive, packages/darwin-mode/src/contrastive
A change to existing behaviour in packages/darwin-mode/dist/contrastive, packages/darwin-mode/src/contrastive — 3 commits (1 fix), 40 files: “feat(darwin): bounded contrastive development API v0.11.0”; “fix(darwin): enforce total contrastive deadline and benchmark integration”; “feat(darwin): independent contrastive confirmation with public pilot evidence”.
packages/darwin-mode/dist/contrastive, packages/darwin-mode/src/contrastive · medium confidence · unverified
3 commits (1 fix) modifying packages/harness
A change to existing behaviour in packages/harness — 3 commits (1 fix), 23 files: “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “feat(harness): receipt chain export/import/merge for cross-system audit interoperability (\#85)”; “feat(harness): ADR-044 Algorithmic Agent Harness control plane (\#6)” (refs \#482, \#85, \#6).
packages/harness · medium confidence · unverified
3 commits (1 fix) modifying packages/workspace-probe
A change to existing behaviour in packages/workspace-probe — 3 commits (1 fix), 9 files: “feat(workspace-probe): CLI (readout/probe/grade-mutation) + runnable demo (0.1.1) (\#136)”; “feat(workspace-probe): @metaharness/workspace-probe — eval + Darwin-Mode bridge for workspace-lens (ADR-239) (\#131)”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)” (refs \#136, \#131, \#482).
packages/workspace-probe · medium confidence · unverified
35 commits (1 fix) modifying packages/darwin-mode/bench/experiments
A change to existing behaviour in packages/darwin-mode/bench/experiments — 35 commits (1 fix), 35 files: “feat(darwin): git-apply patch primitive + whole-file\>raw-diff finding — ADR-098 step-3 prep (ADR-124)”; “feat(darwin): real-LLM evaluation proof-of-concept (ADR-107, toward ADR-098)”; “feat(darwin): crossover ablation refutes ADR-114's mechanism claim — the archive does the work (ADR-115)”; and 32 more.
packages/darwin-mode/bench/experiments · medium confidence · unverified
4 commits (0 fixes) modifying packages/darwin-mode/bench/gaia
A change to existing behaviour in packages/darwin-mode/bench/gaia — 4 commits, 34 files: “feat(orchestration): learned adaptive router vs best static — the capstone "smarter orchestration" test (FRAMES)”; “feat(scaffolding): agent-scaffolding intelligence ablation on FRAMES (cheap models) — scaffolds backfire; only Self-Consistency lifts (not significant)”; “docs(bench): agentic-board integration adapters + roadmap (workflow w9811j59k, 6 boards)”.
packages/darwin-mode/bench/gaia · medium confidence · unverified
4 commits (0 fixes) modifying packages/darwin-mode/bench/terminal-bench
A change to existing behaviour in packages/darwin-mode/bench/terminal-bench — 4 commits, 10 files: “feat(darwin): Terminal-Bench adapter + hardest-first GCP board — eval-validated, hardest probe 1/6 @ $0.09 (§55)”; “feat(terminal-bench): real resolve curve via hardest-first ladder — deepseek 21.2% full-80, Sonnet-4.5 25% hard-band ceiling (LEARNINGS §59)”; “docs(bench): agentic-board integration adapters + roadmap (workflow w9811j59k, 6 boards)”.
packages/darwin-mode/bench/terminal-bench · medium confidence · unverified
4 commits (0 fixes) modifying packages/evals-math
A change to existing behaviour in packages/evals-math — 4 commits, 25 files: “feat(evals-math): LIVE GSM8K wiring — a fast, $0 real-compounding domain testbed (ADR-237) (\#115)”; “feat(evals-math): @metaharness/evals-math — flywheel domain adapter (multi-vertical scale-up) (\#95)”; “docs(evals-math): GSM8K live run evidence — HONEST NULL on a 2nd real domain, $0 (ADR-237 §5) (\#116)”; and 1 more (refs \#115, \#95, \#116, \#176).
packages/evals-math · medium confidence · unverified
4 commits (0 fixes) modifying python
A change to existing behaviour in python — 4 commits, 2 files: “feat(darwin): Terminal-Bench adapter + hardest-first GCP board — eval-validated, hardest probe 1/6 @ $0.09 (§55)”; “feat(terminal-bench): real resolve curve via hardest-first ladder — deepseek 21.2% full-80, Sonnet-4.5 25% hard-band ceiling (LEARNINGS §59)”; “feat: add ChatGPT ARC AGI 3 metaharness (\#218)” (refs \#218).
python · medium confidence · unverified
4 commits (1 fix) modifying packages/horizon
A change to existing behaviour in packages/horizon — 4 commits (1 fix), 17 files: “feat: @metaharness/avo — governed autonomous variation runtime (ADR-251) + live provider validation (\#213)”; “feat: kimi-k3 harness, C-parity Rust/WASM K3 engine, AgentRadio + NOOA clones, darwin/flywheel optimizations — all operational, replay-verified (\#170)”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; and 1 more (refs \#213, \#170, \#482, \#233).
packages/horizon · medium confidence · unverified
4 commits (1 fix) modifying packages/vertical-trading
A change to existing behaviour in packages/vertical-trading — 4 commits (1 fix), 6 files: “feat(iter-15): @ruflo/vertical-base + @ruflo/vertical-trading + external-template loader”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; and 1 more (refs \#482).
packages/vertical-trading · medium confidence · unverified
4 commits (1 fix) modifying packages/weight-eft
A change to existing behaviour in packages/weight-eft — 4 commits (1 fix), 21 files: “dream(darwin-evolution): \#292 reward-hack monitor never scanned tool-response content (evaluated) (\#293)”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “feat(weight-eft): ship @metaharness/weight-eft evolutionary fine-tuning + wire into metaharness (ADR-198)”; and 1 more (refs \#292, \#293, \#482).
packages/weight-eft · medium confidence · unverified
4 commits (2 fixes) modifying packages/jujutsu
A change to existing behaviour in packages/jujutsu — 4 commits (2 fixs), 18 files: “chore(pkg): 0.2.8 hygiene — publish prep, jujutsu drift reconcile, aws-finops (\#71)”; “feat(jujutsu): add @metaharness/jujutsu — version-control-for-agents + agenticow dual-state bridge (ADR-202) (\#56)”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; and 1 more (refs \#71, \#56, \#482).
packages/jujutsu · medium confidence · unverified
4 commits (2 fixes) modifying packages/workspace-lens
A change to existing behaviour in packages/workspace-lens — 4 commits (2 fixs), 14 files: “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “fix: close ADR-index drift, rebuild stale darwin-mode dist, correct README/package.json claims (\#141)”; “feat(workspace-lens): fromUrl + fromRegistry — load a fitted lens over HTTP (0.1.1) (\#132)”; and 1 more (refs \#482, \#141, \#132, \#129).
packages/workspace-lens · medium confidence · unverified
5 commits (0 fixes) modifying apps/web-ui/src/generated
A change to existing behaviour in apps/web-ui/src/generated — 5 commits, 1 file: “feat(template): iter 80 (MILESTONE) — vertical:education, 17th vertical”; “feat(templates): canonical quick-start catalog — 16 verticals, gallery UI, Rust crate, CLI --list”; “feat(template): iter 96 — vertical:gaming, 19th vertical”; and 2 more.
apps/web-ui/src/generated · medium confidence · unverified
5 commits (1 fix) modifying apps/web-ui/scripts
A change to existing behaviour in apps/web-ui/scripts — 5 commits (1 fix), 3 files: “feat(web-ui): social link-preview card + surface score/router in hero”; “chore(studio): benchmark, bundle-split, gen:templates script, README overhaul”; “feat(studio): Repo → Harness importer + in-browser Verify + Studio reframe (ADR-023/024)”; and 2 more (refs \#58).
apps/web-ui/scripts · medium confidence · unverified
5 commits (1 fix) modifying crates/template-catalog
A change to existing behaviour in crates/template-catalog — 5 commits (1 fix), 1 file: “fix(template-catalog): iter 85 — Rust assertion 16 → 17 templates”; “feat(templates): canonical quick-start catalog — 16 verticals, gallery UI, Rust crate, CLI --list”; “feat(template): iter 96 — vertical:gaming, 19th vertical”; and 2 more.
crates/template-catalog · medium confidence · unverified
5 commits (1 fix) modifying packages/darwin-mode/bench/security/fixtures
A change to existing behaviour in packages/darwin-mode/bench/security/fixtures — 5 commits (1 fix), 39 files: “fix(review): deep-review swarm fixes — honest benchmarks + correctness across the PR”; “feat(darwin): real property fuzzer for the invariant loop (ADR-155 Addendum B Phase 2)”; “feat(darwin): real Semgrep oracle — first real analyzer (ADR-155 Addendum A Phase 2)”; and 2 more.
packages/darwin-mode/bench/security/fixtures · medium confidence · unverified
5 commits (2 fixes) modifying packages/vertical-base
A change to existing behaviour in packages/vertical-base — 5 commits (2 fixs), 5 files: “fix(publish): iter 63 — @ruflo/vertical-base ships README + tighten pack-contents”; “feat(iter-15): @ruflo/vertical-base + @ruflo/vertical-trading + external-template loader”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; and 2 more (refs \#482).
packages/vertical-base · medium confidence · unverified
5 commits (2 fixes) modifying submissions/swe-bench-lite
A change to existing behaviour in submissions/swe-bench-lite — 5 commits (2 fixs), 1066 files: “submission(lite): fix accounting — all\_preds is 300 (trailing-newline artifact, not a drop); 35 empty-patch give-ups correctly listed in no\_generation (empties have no eval log by definition); 154/300=51.33% unchanged”; “feat(swebench): assemble acceptance-ready SWE-bench Lite leaderboard submission”; “submission(swe-bench-lite): package Darwin GLM→Opus cascade 51.33% (154/300) — all\_preds + results.json + metadata; cost-Pareto framing ($0.267/inst, conformant, replicated). Ready for swe-bench/experiments PR pending framing sign-off”; and 1 more.
submissions/swe-bench-lite · medium confidence · unverified
50 commits (8 fixes) modifying packages/darwin-mode/src
A change to existing behaviour in packages/darwin-mode/src — 50 commits (8 fixs), 44 files: “feat(darwin): implement Darwin Shield defensive zero-day harness (ADR-155)”; “feat(darwin): wire the ADR-076 graded benchmark gate into evolve() (opt-in)”; “feat(darwin): Tier-2 agent-executing sandbox — real surface code runs (ADR-106)”; and 47 more (refs \#260, \#213, \#465, \#170, \#344, …).
packages/darwin-mode/src · medium confidence · unverified
6 commits (0 fixes) modifying packages/bench/src
A change to existing behaviour in packages/bench/src — 6 commits, 4 files: “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “feat(iter-39): cross-host config-gen benchmark + CLI + 5 tests”; “Prime Agent integration: ADR-241/242 implemented — 10th host, kernel session+autonomous primitives (Rust+TS/wasm), RefineMutator (\#169)”; and 3 more (refs \#169).
packages/bench/src · medium confidence · unverified
6 commits (0 fixes) modifying packages/darwin-mode/bench/cve-bench
A change to existing behaviour in packages/darwin-mode/bench/cve-bench — 6 commits, 32 files: “feat(cve-bench): GCP fleet validation campaign — dataset 4→5 (honest n)”; “feat(cve-bench): era-pinned env builds + SWE-bench recipe reuse — lift §42 yield (ADR-200) (\#54)”; “feat(cve-bench): n=5 dataset + cascade results; update cveSecurity tab”; and 3 more (refs \#54).
packages/darwin-mode/bench/cve-bench · medium confidence · unverified
6 commits (1 fix) modifying packages/agent-harness-generator-lib
A change to existing behaviour in packages/agent-harness-generator-lib — 6 commits (1 fix), 4 files: “feat(packages): iter 116 — dual-package split (mintagent CLI + @ruvnet/agent-harness-generator lib)”; “refactor(brand): iter 118 — mintagent → openharness rename + "Open Harness Studio" UI + Studio SEO”; “feat(brand)+docs: iter 124 — publish \metaharness@0.1.0\ + ADR-035 (product naming)”; and 3 more (refs \#490).
packages/agent-harness-generator-lib · medium confidence · unverified
6 commits (2 fixes) modifying crates/poker-darwin
A change to existing behaviour in crates/poker-darwin — 6 commits (2 fixs), 19 files: “feat(poker-darwin): CFR/CFR+ poker solver + non-stationary Darwin (Chebyshev schedules) (\#49)”; “fix(ci): green the Rust workspace gates (fmt/clippy/test/doc) (\#51)”; “docs(poker-darwin): extend holdem convergence curve to 25k (exploit 0.002215) + runtime note”; and 3 more (refs \#49, \#51, \#60, \#61).
crates/poker-darwin · medium confidence · unverified
6 commits (2 fixes) modifying packages/host-copilot
A change to existing behaviour in packages/host-copilot — 6 commits (2 fixs), 5 files: “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “review(hosts): host capability coverage + live-LLM + real-install verification (ADR-044/045/046) (\#7)”; and 3 more (refs \#482, \#7, \#490, \#331, \#332).
packages/host-copilot · medium confidence · unverified
6 commits (2 fixes) modifying packages/host-opencode
A change to existing behaviour in packages/host-opencode — 6 commits (2 fixs), 5 files: “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “review(hosts): host capability coverage + live-LLM + real-install verification (ADR-044/045/046) (\#7)”; and 3 more (refs \#482, \#7, \#490, \#331, \#332).
packages/host-opencode · medium confidence · unverified
6 commits (2 fixes) modifying packages/router/src
A change to existing behaviour in packages/router/src — 6 commits (2 fixs), 4 files: “feat(router): native FastGRNN backend via @ruvector/tiny-dancer (ADR-043)”; “fix(router): native backend unavailable in real Node ESM (CJS default-unwrap)”; “fix: close ADR-index drift, rebuild stale darwin-mode dist, correct README/package.json claims (\#141)”; and 3 more (refs \#141, \#176).
packages/router/src · medium confidence · unverified
6 commits (2 fixes) modifying packages/sdk
A change to existing behaviour in packages/sdk — 6 commits (2 fixs), 6 files: “feat(iter-14): @ruflo/sdk convenience helpers + browser WASM smoke + preflight + RELEASE runbook”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “fix: integrate Dream changes with security, transport and upgrade regressions covered”; and 3 more (refs \#482, \#418).
packages/sdk · medium confidence · unverified
7 commits (0 fixes) modifying packages/darwin-mode/bench/security
A change to existing behaviour in packages/darwin-mode/bench/security — 7 commits, 6 files: “feat(darwin): implement Darwin Shield defensive zero-day harness (ADR-155)”; “feat(darwin): real Semgrep DRIVES the evolutionary loop (ADR-155 Addendum A Phase 2 capstone)”; “feat(darwin): ablation + hard-corpus stress for Darwin Shield (ADR-155)”; and 4 more.
packages/darwin-mode/bench/security · medium confidence · unverified
7 commits (1 fix) modifying packages/avo/src
A change to existing behaviour in packages/avo/src — 7 commits (1 fix), 17 files: “feat: @metaharness/avo — governed autonomous variation runtime (ADR-251) + live provider validation (\#213)”; “feat(avo): flywheel-gate receipt contract — submit AVO runs to meta-llm's gate (ADR-271) (\#221)”; “feat(core): integrate @metaharness/avo as a metaharness dependency + \metaharness avo\ subcommand (\#226)”; and 4 more (refs \#213, \#221, \#226, \#220, \#217, …).
packages/avo/src · medium confidence · unverified
7 commits (1 fix) modifying packages/host-github-actions
A change to existing behaviour in packages/host-github-actions — 7 commits (1 fix), 6 files: “dream(host-adapters): \#211 close github-actions shell/YAML injection sibling of \#188 (\#212)”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “feat(host-github-actions): port \#480 baseline-vs-candidate policy eval recipe (\#483)”; and 4 more (refs \#211, \#188, \#212, \#480, \#483, …).
packages/host-github-actions · medium confidence · unverified
7 commits (1 fix) modifying packages/host-rvm/src
A change to existing behaviour in packages/host-rvm/src — 7 commits (1 fix), 1 file: “dream(host-adapters): fix host-rvm install-rvm.sh shell+comment injection RCE (evaluated) (\#224)”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “feat(iter-18): harness secrets cmd + GCP validation script + CI/host-rvm fixes”; and 4 more (refs \#224, \#7, \#331, \#332).
packages/host-rvm/src · medium confidence · unverified
8 commits (0 fixes) modifying examples
A change to existing behaviour in examples — 8 commits, 12 files: “docs(examples): iter 129 — 6 examples updated to track \npx metaharness@latest\ (parallel workflow proof)”; “feat(iter-9): federation transport + harness federate + intel pipeline + renovate + examples”; “feat(examples): iter 82 — runnable demo for vertical:education”; and 5 more.
examples · medium confidence · unverified
8 commits (1 fix) modifying apps/web-ui
A change to existing behaviour in apps/web-ui — 8 commits (1 fix), 8 files: “refactor(brand): iter 118 — mintagent → openharness rename + "Open Harness Studio" UI + Studio SEO”; “feat(embeddings): optional in-browser MiniLM semantic engine via Transformers.js (ADR-025)”; “feat(web-ui): social link-preview card + surface score/router in hero”; and 5 more (refs \#169, \#58).
apps/web-ui · medium confidence · unverified
9 commits (0 fixes) modifying packages/darwin-mode/bench/ruvector
A change to existing behaviour in packages/darwin-mode/bench/ruvector — 9 commits, 19 files: “feat(adr-201): H3 empirical verdict — kHop-graph-expansion ≡ dense for ONNX+Wikipedia (structural null)”; “feat(cheap-vs-frontier): ruvector SemanticRouter cost-Pareto pilot (ADR-201 H5) — NOT SUPPORTED on FRAMES”; “feat(adr-201): real OpenRouter account-meter budget gate + corrected H1 pilot config”; and 4 more.
packages/darwin-mode/bench/ruvector · medium confidence · unverified
9 commits (1 fix) modifying crates/kernel/src
A change to existing behaviour in crates/kernel/src — 9 commits (1 fix), 14 files: “feat(iter-1): scaffold Cargo + npm workspaces, kernel stubs, CI/publish/security pipeline”; “feat(iter-8): harness sign/verify/doctor CLI + MCP tool registry + 6 package READMEs + GCP setup script”; “feat(iter-10): MCP tool dispatch + cost tracking + AST rename + tarball builder + integration smoke”; and 6 more (refs \#169).
crates/kernel/src · medium confidence · unverified
9 commits (2 fixes) modifying packages/host-codex
A change to existing behaviour in packages/host-codex — 9 commits (2 fixs), 15 files: “feat(iter-8): harness sign/verify/doctor CLI + MCP tool registry + 6 package READMEs + GCP setup script”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; and 6 more (refs \#482, \#7, \#490, \#299, \#300, …).
packages/host-codex · medium confidence · unverified
9 commits (4 fixes) modifying packages/host-openclaw
A change to existing behaviour in packages/host-openclaw — 9 commits (4 fixs), 5 files: “dream(host-adapters): fix host-openclaw YAML/shell injection via spec.name (evaluated) (\#246)”; “feat(iter-11): add @ruflo/host-openclaw — 5th host adapter”; “refactor!: rename @ruflo/\* packages to @metaharness/\* scope”; and 6 more (refs \#246, \#482, \#7, \#490, \#331, …).
packages/host-openclaw · medium confidence · unverified
Fixes
1 commit (1 fix) fixing bench
A fix in bench — 1 commit (1 fix), 1 file: “fix: integrate Dream changes with security, transport and upgrade regressions covered”.
bench · medium confidence · unverified
1 commit (1 fix) fixing packages/create-agent-harness/assets
A fix in packages/create-agent-harness/assets — 1 commit (1 fix), 1 file: “fix(generator): make installed harness journeys executable and test policy fitness (\#439)” (refs \#439).
packages/create-agent-harness/assets · medium confidence · unverified
1 commit (1 fix) fixing packages/router/bench
A fix in packages/router/bench — 1 commit (1 fix), 3 files: “fix: close ADR-index drift, rebuild stale darwin-mode dist, correct README/package.json claims (\#141)” (refs \#141).
packages/router/bench · medium confidence · unverified
3 commits (2 fixes) fixing packages/host-prime-agent
A fix in packages/host-prime-agent — 3 commits (2 fixs), 7 files: “Prime Agent integration: ADR-241/242 implemented — 10th host, kernel session+autonomous primitives (Rust+TS/wasm), RefineMutator (\#169)”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “fix(agntcy, agent-harness-generator): install without the Buf registry; scaffold across filesystems; ADR-356 residuals (\#490)” (refs \#169, \#482, \#490).
packages/host-prime-agent · medium confidence · unverified
5 commits (3 fixes) fixing packages/arc-agi-3-chatgpt
A fix in packages/arc-agi-3-chatgpt — 5 commits (3 fixs), 29 files: “fix: port \#381's dependency-advisory fix to unblock npm-audit CI”; “feat(arc): add governed AVO harness, ChatGPT adapter, and benchmarks (\#220)”; “fix(security): update validation and remove unused vulnerable arena dependency (\#441)”; and 2 more (refs \#381, \#220, \#441, \#269, \#218).
packages/arc-agi-3-chatgpt · medium confidence · unverified
8 commits (5 fixes) fixing packages/agntcy
A fix in packages/agntcy — 8 commits (5 fixs), 27 files: “fix(agntcy): exclude nested \_\tests\\_ dirs from tsc build output”; “seo(ADR-356): metadata, READMEs and quickstart fixes for 69 @metaharness packages (\#482)”; “ADR-240: AGNTCY identity, OASF export, observability, CASA compiler (\#155)”; and 5 more (refs \#482, \#155, \#490, \#156, \#158, …).
packages/agntcy · medium confidence · unverified
Test coverage
11 commits adding/updating tests in packages/projects/\_\tests\\; 12 commits adding/updating tests in packages/kernel-js/\\tests\\; 16 commits adding/updating tests in packages/darwin-mode/\\tests\\/security; 2 commits adding/updating tests in \\tests\\/browser-smoke; 2 commits adding/updating tests in \\tests\\/integration; 2 commits adding/updating tests in packages/darwin-mode/\\tests\\/bench; 31 commits adding/updating tests in packages/bench/\\tests\\; 34 commits adding/updating tests in packages/darwin-mode/\\tests\\; 4 commits adding/updating tests in packages/darwin-mode/\\tests\\/perf; 5 commits adding/updating tests in packages/router/\\tests\\; 6 commits adding/updating tests in packages/avo/\\tests\\; 78 commits adding/updating tests in \\tests\\; 8 commits adding/updating tests in apps/web-ui/e2e; 8 commits adding/updating tests in packages/darwin-mode/\\tests\\/e2e; 8 commits adding/updating tests in packages/host-rvm/\\tests\\; 81 commits adding/updating tests in packages/create-agent-harness/\\tests\\_.
Dependencies
214 commits updating dependencies (102 manifests)
A dependency / build maintenance change in (dependencies) — 214 commits (51 fixs), 102 files: “feat(darwin): implement Darwin Shield defensive zero-day harness (ADR-155)”; “feat(metaharness): add @metaharness/darwin as a dependency + bump scaffold version to ^0.2.2”; “chore(release): metaharness 0.1.8 — uniform publishConfig in all emitted templates”; and 211 more (refs \#118, \#145, \#167, \#442, \#229, …).
(dependencies) · high confidence · unverified
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
How this codebase got here
Baseline
- First survey — no prior run to compare against. CAI 50.3.
Lenses
- Code Health 63.2
- Architecture 72.1
- Maturity 96.2
- Readiness 51.3
- Security 71.1
- Event Sourcing 100.0
- Accessibility 37.3
- Performance 72.9
Changes since last survey
- 300 commits — 230 feature/other, 70 fixes
By area
- packages/darwin-mode — 50 commits
- packages/create-agent-harness — 34 commits
- (root) — 28 commits
- docs/research — 26 commits
- docs/adrs — 21 commits
- docs/dream-cycle — 20 commits
- (repo) — 18 commits
- .github/workflows — 17 commits
- packages/flywheel — 16 commits
- packages/agntcy — 7 commits
- apps/web-ui — 5 commits
- packages/avo — 5 commits
- packages/kernel-js — 5 commits
- packages/redblue — 5 commits
- docs/assets — 4 commits
- packages/arc-agi-3-chatgpt — 4 commits
- docs/releases — 3 commits
- packages/evals-math — 3 commits
- packages/jujutsu — 3 commits
- services/apicompletions — 3 commits
Notable commits
- fix: Merge Dream integration, security fixes, and 0.4.18 release preparation (#426)
- fix: Merge pull request #251 from ruvnet/fix/meta-proxy-stale-owner
- fix: Merge pull request #498 from ruvnet/fix/jujutsu-agenticow-pin
- fix: Merge retort metaharness iteration-2 into main (timeout fix + memory factor; beyond-SOTA not reached)
- fix: chore(redblue): release 0.1.5 — indirect_prompt_injection structured-field bypass fix
- fix: chore(release): @metaharness/darwin 0.9.1 — ships #167 mock-bench sandbox fix
- fix: dream(flywheel-promotion): #340 fix flywheel graph RangeError on negative-primary lift points (evaluated) (#341)
- fix: dream(host-adapters): fix hermes YAML mapping-key injection, all 3 codegen paths (evaluated) (#188)
- fix: dream(host-adapters): fix host-openclaw YAML/shell injection via spec.name (evaluated) (#246)
- fix: dream(host-adapters): fix host-rvm install-rvm.sh shell+comment injection RCE (evaluated) (#224)
- fix: fix(agntcy): exclude nested tests dirs from tsc build output
- fix: fix(agntcy): schema_version mismatch was the real root cause, not a validator bug (#157)
- fix: fix(agntcy, agent-harness-generator): install without the Buf registry; scaffold across filesystems; ADR-356 residuals (#490)
- fix: fix(arc-agi-3-chatgpt): give fs-heavy store tests explicit timeouts (#269)
- fix: fix(avo)!: isolate agent context — untrusted agent could mutate live state and buy promotion (#214)
- fix: fix(ci): make publish.yml idempotent — skip already-published versions (#153)
- fix: fix(ci): rustfmt + cargo-deny advisories (unblock PR #60) (#61)
- fix: fix(ci): unbreak the Node matrix — path-guard false positives + cost-breaker e2e timeout (#75)
- fix: fix(cli): propagate main() exit code in bin.js — failed commands now exit non-zero (#73) (#78)
- fix: fix(create-agent-harness): bump @metaharness/darwin pin ^0.2.2 → ^0.8.0 (#142) (#150)
- …and 280 more
Architecture
- 0 containers · 1 bounded contexts · 0 dependency edges (baseline)
Written by cca.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.
Survey your own repository
ruvnet/metaharness was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.
About this page
- The score is its most recent published measurement, taken on 17 November 2026 at a pinned commit. It is not a live figure and does not change until the project is measured again.
- Measured at commit 1dc8f009771c763c9b37d04d17e7bd8a9ec7feba — the exact code this score is about.
- Scored under rubric-2026.11.4 — the same rubric and the same method as every other entry in this index.
- Measured by Canine Code Analyzer (cca.canine.dev) using cca measure-2026.11.84+preprod-d645ca9bbedd.