github.com/ruvnet/ruflo...
github.com/ruvnet/ruflo...
AgentDB dependency floor raised from ^3.0.0-alpha.17 to ^3.0.0-alpha.20 across memory, cli, root, and ruflo packages; AgentDB alpha.20 replaces @xenova/transformers with @huggingface/transformers, upgrades @ruvector/rvf from 0.1.9 to 0.2.3, and moves ruvector to 0.2.x; Search scores…
AgentDB dependency floor raised from ^3.0.0-alpha.17 to ^3.0.0-alpha.20 across memory, cli, root, and ruflo packages; AgentDB alpha.20 replaces @xenova/transformers with @huggingface/transformers, upgrades @ruvector/rvf from 0.1.9 to 0.2.3, and moves ruvector to 0.2.x; Search scores…
Upgrading closes the entry point. It does not remove instructions already written into memory. Those get retrieved later and shape how agents answer other users' tasks.
Upgrading closes the entry point. It does not remove instructions already written into memory. Those get retrieved later and shape how agents answer other users' tasks.
- 클로드 코드에 **100개 이상의 특화 에이전트** 를 조율하는 능력을 부여하는 플랫폼으로, 기존 Claude Flow에서 리브랜딩
- `npx ruvflo init` 한 줄이면 에이전트가 **스웜으로 자기 조직화** 하고, 작업마다 학습하며, 세션 간 기억을 유지하고, 다른 머신의 에이전트와 보안 통신을 시작
- **HNSW 벡터 메모리**(AgentDB) 기반으로 brute-force 대비 150배~12,500배 빠른 검색, **SONA […]
- 클로드 코드에 **100개 이상의 특화 에이전트** 를 조율하는 능력을 부여하는 플랫폼으로, 기존 Claude Flow에서 리브랜딩
- `npx ruvflo init` 한 줄이면 에이전트가 **스웜으로 자기 조직화** 하고, 작업마다 학습하며, 세션 간 기억을 유지하고, 다른 머신의 에이전트와 보안 통신을 시작
- **HNSW 벡터 메모리**(AgentDB) 기반으로 brute-force 대비 150배~12,500배 빠른 검색, **SONA […]
Claude → DeepSeek Flash: 99.5% cheaper.
GitHub Actions → free cron.
MCP server → no API layer needed.
Knowledge API for AI agents. 41 sources, MCP-ready.
Try it: agentdb.dev
#MCP #DeepSeek #BuildInPublic
Claude → DeepSeek Flash: 99.5% cheaper.
GitHub Actions → free cron.
MCP server → no API layer needed.
Knowledge API for AI agents. 41 sources, MCP-ready.
Try it: agentdb.dev
#MCP #DeepSeek #BuildInPublic
60× compression. 44 sources. Updated 3x a week. Any agent can query it via API.
agentdb-production-9ba0.up.railway.app
#buildinpublic #AI
60× compression. 44 sources. Updated 3x a week. Any agent can query it via API.
agentdb-production-9ba0.up.railway.app
#buildinpublic #AI
Ruflo is a multi-agent AI orchestration platform for Claude Code, enabling 100+ autonomous agents to swarm, learn, memory-share, and securely federate across machines with zero-trust federation, rich memory (AgentDB/HNSW), and a web UI for multi-model chat and GOAP planning. It (1/2)
Ruflo is a multi-agent AI orchestration platform for Claude Code, enabling 100+ autonomous agents to swarm, learn, memory-share, and securely federate across machines with zero-trust federation, rich memory (AgentDB/HNSW), and a web UI for multi-model chat and GOAP planning. It (1/2)
I still managed to ship in 3 days, which would typically take 6-8 weeks.
The real Agentic QE Engineering story -
www.linkedin.com/pulse/leadin...
I still managed to ship in 3 days, which would typically take 6-8 weeks.
The real Agentic QE Engineering story -
www.linkedin.com/pulse/leadin...
✔️ thedataexchange.media/agentdb-luke...
✔️ thedataexchange.media/agentdb-luke...
thedataexchange.media/agentdb-luke...
thedataexchange.media/agentdb-luke...