Ember
Production-ready AI agent framework built for enterprise deployments. Create intelligent, autonomous agents that reason, plan, and execute with safety and security at every step.
Built for Production AI Agents
Ember implements the GAME methodology, providing a complete foundation for building intelligent, autonomous agents.
GAME Architecture
Built on the Goals, Actions, Memories, Environment methodology. Define weighted goals, execute actions, maintain multi-tier memory, and enforce environmental constraints.
Execution Loop
Sophisticated plan → act → reflect → finalize loop with configurable max steps. Self-evaluation ensures quality responses before finalizing.
Multi-Tier Memory System
Ephemeral scratchpad for current tasks, sliding window session memory, and long-term vector storage with Qdrant for semantic search, retrieval, and continuous learning.
Native Tool Calling
Rich metadata with auto-generated JSON schemas from type hints. Built-in HTTP client with domain allowlists, MCP client support, and extensible @tool decorator for custom tools.
Dynamic Plugin System
Drop-in tool plugins with automatic discovery from entry points and filesystem. Per-tool configuration with Pydantic validation and zero core modifications required.
Security Guardrails
Jailbreak detection with pattern-based prompt injection prevention. Granular secure tool controls. HTTP domain and method allowlisting, scope enforcement, and automatic secret redaction from logs.
Multi-Protocol API Access
Expose your agents via HTTP with JSON-RPC 2.0, REST API, and WebSocket support. Built-in authentication, rate limiting, and webhook delivery.
JSON-RPC 2.0
Standard protocol support
REST API
RESTful endpoints
WebSocket
Real-time streaming
Admin API
Management endpoints
Flexible Model Support
OpenAI
GPT-5, GPT-4, GPT-4o, GPT-3.5-turbo
OpenRouter
100+ models via unified API
Extensible
Add additional providers via provider plugins
Production-Ready Deployment
Deploy Ember agents anywhere with flexible options for containers, orchestration, and microVMs.
Docker
Container-ready with Docker Compose support
Kubernetes
Production-grade orchestration
Firecracker
Lightweight microVM deployment
Multi-Agent Fleet
Scale to hundreds or even thousands of agents
Comprehensive Audit Logging
Structured JSON logging with 12+ event types, automatic sensitive data redaction, request correlation, and performance metrics for complete visibility.
Native Lens Integration
Ember natively integrates with Veilfire Lens for real-time AI safety monitoring, policy enforcement, and cryptographic audit logging. Deploy with confidence knowing every agent action is governed and auditable.