HN 표시: EDDI – 에이전트 로직이 코드가 아닌 JSON에 있는 다중 에이전트 AI 엔진

E.D.D.I (Enhanced Dialog Driven Interface) is a production-grade, config-driven multi-agent orchestration middleware for conversational AI. It coordinates users, AI agents, and business systems through intelligent routing, persistent memory, and API orchestration — without writing code. Built with Java 25 and Quarkus. Ships as a Red Hat-certified Docker image. Native support for MCP (Model Context Protocol), A2A (Agent-to-Agent), Slack, OpenAPI, and OAuth 2.0. Latest version: 6.0.1 · Website · Documentation · License: Apache 2.0 - 🏁 Quick Start - 💡 Why EDDI? - 📸 See It In Action - ✨ Features - 🧩 Quarkus SDK - 📖 Documentation - 📋 Compliance & Privacy - 🏗️ Development - 🤝 Contributing - 🔒 Security - 📜 Code of Conduct The fastest way to get EDDI running is the one-command installer. It sets up EDDI + your choice of database via Docker Compose, deploys the Agent Father starter agent, and walks you through creating your first AI agent. Linux / macOS / WSL2: curl -fsSL https://raw.githubusercontent.com/labsai/EDDI/main/install.sh | bash Windows (PowerShell): iwr -useb https://raw.githubusercontent.com/labsai/EDDI/main/install.ps1 | iex Note: If your Antivirus blocks this command as "malicious content", securely download and run it instead: Invoke-WebRequest -Uri "https://raw.githubusercontent.com/labsai/EDDI/main/install.ps1" -OutFile "install.ps1" Unblock-File .\install.ps1 .\install.ps1 Requires Docker. The wizard auto-generates a unique vault encryption key for secret management. 🔧 Installer options bash install.sh --defaults # All defaults, no prompts bash install.sh --db=postgres --with-auth # PostgreSQL + Keycloak bash install.sh --full # Everything enabled (DB + auth + monitoring) bash install.sh --local # Build Docker image from local source The --local flag is for contributors testing pre-release builds: ./mvnw package -DskipTests # Build the Java app bash install.sh --local # Build Docker image + start containers The installer creates an eddi CLI wrapper that makes updating easy: eddi update This pulls the latest Docker image from the registry and restarts the containers. It works even when the same tag (e.g. latest ) was re-published — Docker always checks the remote digest for changes. eddi command not found? The CLI lives at~/.eddi/eddi (Linux/macOS) or~/.eddi/eddi.cmd (Windows). Either restart your terminal so the PATH takes effect, or use the full path:# Linux / macOS ~/.eddi/eddi update # Windows (PowerShell) & "$HOME\.eddi\eddi.cmd" update Manual update (without the CLI) If the eddi CLI isn't available, run the equivalent docker commands from your install directory (~/.eddi by default): cd ~/.eddi docker compose --env-file .env -f docker-compose.yml pull docker compose --env-file .env -f docker-compose.yml up -d Adjust the -f flags to match your setup (e.g. add -f docker-compose.auth.yml if using Keycloak). If you prefer manual control over Docker Compose: # Default (EDDI + MongoDB) docker compose up # PostgreSQL instead of MongoDB EDDI_DATASTORE_TYPE=postgres docker compose -f docker-compose.yml -f docker-compose.postgres.yml up # With Keycloak authentication docker compose -f docker-compose.yml -f docker-compose.auth.yml up # With Prometheus + Grafana monitoring docker compose -f docker-compose.yml -f docker-compose.monitoring.yml up # Full stack (all overlays) docker compose -f docker-compose.yml -f docker-compose.auth.yml \ -f docker-compose.monitoring.yml -f docker-compose.nats.yml up Available compose overlays: docker-compose.auth.yml (Keycloak), docker-compose.monitoring.yml (Prometheus+Grafana), docker-compose.nats.yml (NATS JetStream), docker-compose.postgres.yml / docker-compose.postgres-only.yml , docker-compose.local.yml (build from source). docker pull labsai/eddi # Pull latest from Docker Hub Most multi-agent frameworks (LangGraph, CrewAI, AutoGen) are Python/Node libraries — great for prototyping, hard to govern in production. EDDI approaches from the opposite direction: a deterministic engine built to safely govern non-deterministic AI. | Dimension | Typical Python/Node Frameworks | EDDI | |---|---|---| | Concurrency | GIL or single-threaded event loop | Java 25 Virtual Threads — true OS-level parallelism | | Agent Logic | Embedded in application code | Versioned JSON configurations — update behavior without redeployment | | Security Model | Often relies on sandboxed code execution | No dynamic code execution at all; envelope-encrypted vault, SSRF protection | | Compliance | Requires custom implementation | GDPR, HIPAA, EU AI Act infrastructure built-in | | Audit Trail | Application-level logging | HMAC-SHA256 immutable ledger with cryptographic agent signing | | Deployment | pip/npm + manual infrastructure | One-command Docker install, Kubernetes/OpenShift-ready | "The engine is strict so the AI can be creative." — Project Philosophy More screenshots: LLM Config, Logs, User Memory, Schedules, Agent Detail - 🔀 Intelligent Routing — Direct conversations to different agents based on context, rules, and intent - 🗣️ Group Conversations — Multi-agent debates with 5 built-in discussion styles: Round Table, Peer Review, Devil's Advocate, Delphi, and Debate - 💬 Slack Integration — Deploy agents to Slack channels and run multi-agent debates directly in threads - 🪆 Nested Groups — Compose groups of groups for tournament brackets, red-team vs blue-team, and panel reviews - 👥 Managed Conversations — Intent-based auto-routing with one conversation per user per intent - 🎯 Capability Matching — Discover and route to agents by skill, confidence score, and custom attributes - 🧙 Agent Father — Meta-agent that creates other agents through conversation (ships out of the box) | Category | Providers | |---|---| | Cloud APIs | OpenAI · Anthropic Claude · Google Gemini · Mistral AI | | Enterprise Cloud | Azure OpenAI · Amazon Bedrock · Oracle GenAI · Google Vertex AI | | Self-Hosted | Ollama · Jlama · Hugging Face | | Compatible | Any OpenAI-compatible endpoint (DeepSeek, Cohere, etc.) via baseUrl | EDDI implements open standards — not proprietary APIs: | Standard | Role | What It Enables | |---|---|---| | MCP (Model Context Protocol) | Server (42 tools) + Client | Control EDDI from Claude Desktop, Cursor, or any MCP client. Connect agents to external MCP tool servers | | A2A (Agent-to-Agent Protocol) | Full implementation | Cross-platform agent communication, Agent Cards, and skill discovery | | OpenAPI 3.1 | Native generation + consumption | Auto-generated spec. Paste any OpenAPI spec → get a fully deployed API-calling agent | | OAuth 2.0 / OIDC | Keycloak integration | Authentication, authorization, and multi-tenant isolation | | SSE (Server-Sent Events) | Streaming transport | Real-time chat responses, group discussion feeds, and live log streaming | - 💾 Persistent User Memory — Agents remember facts, preferences, and context across conversations via structured key-value entries with visibility scoping ( global ,agent ,group ) - 🧠 LLM Memory Tools — Built-in tools agents can call to read, write, and search their own persistent memory - 💤 Dream Consolidation — Background memory maintenance: stale entry pruning, contradiction detection, and fact summarization (inspired by Anthropic's research on background memory consolidation) - 🪟 Token-Aware Windowing — Intelligent context packing with model-specific tokenizer support and anchored opening steps - 📝 Rolling Summary — Incremental LLM-powered summarization of older turns with a Conversation Recall Tool for drill-back into compressed history - 🔧 Property Extraction — Config-driven slot-filling with longTerm /conversation /step scoping — EDDI's importance extraction mechanism - 🛡️ Memory Policy (Commit Flags) — Strict write discipline marks failed task output as uncommitted (hidden from LLM context) and injects concise error digests for graceful degradation - 🔄 Conversation State — Full history with undo/redo support - 📦 7 Embedding Providers — OpenAI, Ollama, Azure OpenAI, Mistral, Bedrock, Cohere, Vertex AI - 🗄️ 5 Vector Stores — pgvector, In-Memory, MongoDB Atlas, Elasticsearch, Qdrant - 🌐 httpCall RAG — Zero-infrastructure RAG via any search API (BM25, Elasticsearch, custom) - 📥 REST Ingestion API — Async document ingestion with status tracking | Tool | Description | |---|---| | 🔍 Web Search | DuckDuckGo or Google Custom Search | | 🧮 Calculator | Sandboxed recursive-descent math parser (no eval() , no code injection) | | 🌐 Web Scraper | SSRF-protected content extraction from web pages | | 📄 PDF Reader | SSRF-protected document extraction | | ☁️ Weather · 🕐 DateTime | Real-time data tools | | 📊 Data Formatter · 📝 Text Summarizer | Data transformation tools | | 🔌 HTTP Calls as Tools | Expose your own REST APIs as LLM-callable tools with full security sandboxing | | 🧠 User Memory | Read/write/search persistent user memory | | 🔙 Conversation Recall | Drill back into summarized conversation history | | 📎 Multimodal Attachments | Image, PDF, audio, and video input with MIME-based routing | - 🫀 Heartbeat Triggers — Periodic agent wake-ups at configurable intervals for proactive behavior (inspired by OpenClaw's heartbeat architecture) - ⏲️ Cron Scheduling — Standard cron expressions for timed agent execution - 🔄 Conversation Strategies — persistent (reuse same conversation across fires) ornew (fresh context each time) - 📊 Fire Logging — Complete execution history with status, duration, cost tracking, and retry logic - 🌙 Dream Cycles — Scheduled background memory consolidation with cost ceilings per run - 📉 Cost Optimization — Try cheap/fast models first, escalate to powerful models only when confidence is low - 📊 4 Confidence Strategies — Structured output, heuristic, judge model, or none - 💰 Per-Conversation Budgets — Automatic cost tracking with budget caps and eviction - 🏢 Tenant Cost Ceilings — Monthly cost budgets per tenant with automatic enforcement Security Architecture - 🏦 Secrets Vault — Envelope encryption (PBKDF2 + AES-256) with tenant-scoped DEK/KEK rotation. Never plaintext in DB - 🛡️ SSRF Protection — All tools validate URLs against private IPs, internal hostnames, and non-HTTP schemes before any request - 🔒 Sandboxed Evaluation — Recursive-descent math parser only. No eval() , no script engines, no reflection-based execution - 🔑 OAuth 2.0 / Keycloak — Multi-tenant authentication, authorization, and role-based access control - ✍️ Agent Signing — Ed25519 cryptographic identity per agent; audit entries signed with agent private keys - 🚫 No Dynamic Code Execution — Custom logic runs in external MCP servers, outside the EDDI security perimeter Regulatory Compliance | Regulation | EDDI Support | |---|---| | EU AI Act | Immutable HMAC-SHA256 audit ledger, decision traceability, risk classification guidance | | GDPR | Cascading data erasure (Art. 17), data portability (Art. 15/20), restriction of processing (Art. 18), per-category retention, pseudonymization | | CCPA | Right to delete, right to know, data portability | | HIPAA | Deployment guide, BAA template, LLM provider BAA matrix, session timeout guidance | | International | PIPEDA 🇨🇦 · LGPD 🇧🇷 · APPI 🇯🇵 · POPIA 🇿🇦 · PDPA 🇸🇬🇹🇭🇲🇾 · PIPL 🇨🇳 compatibility documented | - 📜 Audit Ledger — Every agent decision recorded in a write-once, HMAC-secured, append-only ledger - 🔍 Compliance Startup Checks — Advisory warnings on boot for TLS and database encryption gaps - 🗑️ GDPR Orchestration — One-call cascading erasure across 6 stores + audit trail pseudonymization - 📤 Data Portability — Complete user data export (memories, conversations, audit entries) via REST and MCP - 📄 JSON Configs, Not Code — Agent behavior defined in versioned, diffable JSON documents - 🔧 Lifecycle Pipeline — Pluggable task pipeline: Input → Parse → Rules → API/LLM → Output - 📦 Composable Agents — Agents assembled from reusable, version-controlled workflows and extensions - 🧪 Behavior Rules — IF-THEN logic engine for routing, orchestration, and business logic - 📤 Import / Export — Agents portable as ZIP files with automatic secret scrubbing on export - 🔄 Agent Sync — Live instance-to-instance sync with structural matching, content diffing, and selective resource picking — no ZIP intermediary needed - 📝 Prompt Snippets — Reusable, versioned system prompt building blocks available as {{snippets.safety_rules}} - 📎 Content Type Routing — MIME-based behavior rule conditions for multimodal attachment routing - 🐳 One-Command Install — Interactive wizard sets up EDDI + database + starter agent via Docker - ☸️ Kubernetes / OpenShift — Kustomize overlays, Helm charts, HPA, PDB, NetworkPolicy - 📊 Prometheus & Grafana — 50+ Micrometer metrics at /q/metrics (tools, vault, memory, scheduling, conversations) - 🩺 Health Checks — Liveness & readiness probes at /q/health/live and/q/health/ready - 🔄 NATS JetStream — Async event bus for distributed processing - ⚡ Virtual Threads — Java 25 virtual threads for true OS-level concurrency (no Python GIL or Node.js event loop bottleneck) - 🗃️ DB-Agnostic — Choose MongoDB or PostgreSQL; switch with one env var. Single Docker image for both - 🏗️ Red Hat Certified — Container certification with automated preflight checks in CI/CD - 🎨 React 19 Manager — Modern admin dashboard for agent building, testing, deployment, and monitoring - 💬 Chat Widget — Embeddable React chat UI with SSE streaming and Keycloak auth - 🔍 Audit Trail Viewer — Timeline-based compliance and debugging UI - 📋 Logs Panel — Live SSE log streaming + searchable history - 🔑 Secrets Manager — Write-only vault UI with copy-reference support - 🌍 11 Languages — English, German, Spanish, French, Portuguese, Chinese, Japanese, Korean, Arabic (RTL), Hindi, Thai Building a Quarkus app that talks to EDDI? Use the quarkus-eddi extension: io.quarkiverse.eddi quarkus-eddi 6.0.1 @Inject EddiClient eddi; String answer = eddi.chat("my-agent", "Hello!"); Features: Dev Services (auto-starts EDDI in dev mode), fluent API, SSE streaming, @EddiAgent endpoint wiring, @EddiTool MCP bridge. See the quarkus-eddi README for full docs. | Guide | Description | |---|---| | Getting Started | Setup and first steps | | Developer Quickstart | Build your first agent in 5 minutes | | Architecture | Deep dive into EDDI's design and pipeline | | LLM Configuration | Connecting to 12 LLM providers | | Behavior Rules | Configuring agent routing logic | | HTTP Calls | External API integration | | RAG | Knowledge base retrieval setup | | MCP Server | 42 tools for AI-assisted agent management | | A2A Protocol | Agent-to-Agent peer communication | | Slack Integration | Deploy agents to Slack and run group discussions | | Group Conversations | Multi-agent debate orchestration | | User Memory | Cross-conversation fact retention | | Memory Policy | Commit flags and strict write discipline | | Model Cascading | Cost-optimized multi-model routing | | Scheduling & Heartbeats | Cron schedules, heartbeats, dream consolidation | | Agent Sync | Live instance-to-instance sync and upgrade imports | | Import