🌊 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
TypeScript agentic-aiagentic-frameworkagentic-workflowagentsai-agentsai-assistantai-skillsautonomous-agentsclaudecodexdsh-pluginharnessmcp-servermulti-agentmulti-agent-systemsnpmskillsswarmswarm-intelligencetypescript
Commits on main
chore: sync repo dogfood helpers to 3.23.0 + gitignore auto-refresh runtime stamp
a5f86ad ruv committed 4 months ago
feat(memory): nightly vector-DB backup — WAL-safe snapshot, rotation, daemon worker (#2571)
802b75b rUv committed 4 months ago
feat: memory distillation self-learning loop (ADR-174) + page-agent browser intent (ADR-175) (#2570)
cd8a9c0 rUv committed 4 months ago
fix(statusline): real vector count + self-heal missing vector_indexes on init (#2569)
5e29dc1 rUv committed 4 months ago
perf: optimize memory-search hot loop, agenticow label lookup, and npm packaging (#2568)
a1767fc rUv committed 4 months ago
chore(release): 3.21.0 → 3.21.1 (memory-search recall #2558 + init self-learning #2545)
cc354c6 ruv committed 4 months ago
fix(memory): restore keyword+semantic recall in `memory search` (#2558) (#2567)
51085bf rUv committed 4 months ago
fix(init): wire self-learning memory bridge on the npx path (#2545) (#2564)
4622884 rUv committed 4 months ago
chore(release): 3.20.0 → 3.21.0 (agenticow governed loop, ADR-170..173)
a5a48fa ruv committed 4 months ago
feat(agenticow): COW memory substrate + governed distillation loop (ADR-170..173) (#2562)
d759af6 rUv committed 4 months ago
ci: fast lockfile-drift guard — catch pnpm-lock↔manifest mismatch early (#2562) (#2563)
5def666 rUv committed 4 months ago
chore(release): 3.19.0 → 3.20.0 (training flywheel + ruvllm 2.6.0)
d0e77d8 ruv committed 4 months ago
feat(neural): training flywheel + @ruvector/ruvllm 2.6.0 (resume, val-split, checkpoint auto-load) (#2557)
66ba809 rUv committed 4 months ago
chore(release): 3.18.2 → 3.19.0 (native training pipeline routing)
1155b42 ruv committed 4 months ago
feat(neural): route neural train through the native TrainingPipeline (#2549 final follow-up) (#2556)
4bea342 rUv committed 4 months ago
chore(release): 3.18.1 → 3.18.2 (native ruvllm checkpoints end-to-end)
feb6ecb ruv committed 4 months ago
feat(neural): use @ruvector/ruvllm 2.5.7 native checkpoints end-to-end (#2549 follow-up) (#2555)
8547df9 rUv committed 4 months ago
fix(neural): #2549 — status no longer misreports native ruvllm training path as Unavailable (#2554)
9b504d2 rUv committed 4 months ago
chore(release): 3.17.0 → 3.18.0 (review-driven upgrades + GAIA integrity family)
dc82e7f ruv committed 4 months ago
feat: review-driven upgrades — perf, security, SOTA capabilities (6-agent concurrent implementation) (#2547)
d012042 rUv committed 4 months ago
feat(gaia): pre-submission exploit audit + signed attestation (ADR-167) (#2543)
e0d0af1 rUv committed 4 months ago
fix(ci): regenerate v3/pnpm-lock.yaml for the #2540 metaharness pin bumps (#2552)
d07ec82 rUv committed 4 months ago
feat(metaharness): learn + gepa integration — [email protected] / [email protected] (15 MCP tools) (#2540)
ae9bba5 rUv committed 4 months ago
docs(claude): metallm ask-vs-delegate rule + $0.26 delegate floor warning (#2538)
6f89c76 rUv committed 4 months ago
docs(claude): enforce gateway-delegated development via meta-llm dev-bridge (#2535)
0810d3b rUv committed 4 months ago