whereismyai
OmniAgent
OmniAgent is an open-source self-evolving agent framework that implements full-dimensional self-evolution across skills, context, and models.
OmniAgent is an open-source self-evolving agent framework inspired by OpenClaw. It implements full-dimensional self-evolution through proactive memory, skill self-evolution, personalization context, and brainmodel self-evolution. The system uses a multi-layer information stack and online reinforcement learning to continuously update memory, user preferences, and models. Additionally, safety and task success rates are enhanced through specialized architectural modules. It supports multiple LLM backends including OpenAI, Claude, DeepSeek, Ollama, and Gemini. Users interact with it via terminal, web user interface, and mobile channels like Feishu, Discord, and Telegram.
OmniAgent at a glance
- Best for: Software engineer
- Runs on: Web, CLI, self-hosted, open source
- Pricing: Free and open source under GPL-3.0
- Released: April 2026
- Category: AI agent tool
- Website: github.com/YeQing17-2026/OmniAgent
Key features
- Proactive Memory for continuous evolution
- Real-time skill self-evolution
- Adaptive personalized context construction
- BrainModel self-evolution via online reinforcement learning
- Progressive context loading for token optimization
- Dynamic multi-agent task analysis
- Dynamic concurrent tool execution
- Four-layer dynamic security scanning
Use it for
- Define agent personality and rules via bootstrap files
- Write, run, and test code locally with auto-correction
- Execute multi-source web searches and cross-validate knowledge
- Run shell commands with built-in safety control flows
- Manage message routing across Feishu, Discord, and Telegram
Alternatives to OmniAgent
- Omnigent — Manage, secure, and collaborate on diverse AI agents from one central layer.
- EvoAgentX — It automatically builds, evaluates, and evolves multi-agent workflows.
- Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
- Beeuni — Agents retain context and improve skills over time through a controlled review process.
- VoltAgent — Offers open-source flexibility with built-in supervisor agent orchestration and RAG.
- MyAgent — It features built-in permission controls, long-term memory, and skill self-evolution.
- Harness2 — It recursively refines agent execution structures using parallel and sequential search.
- OpenViking — Unifies agent memory, knowledge RAG, and skills into a single directory structure.
- OpenSquilla — Maximizes AI intelligence density while minimizing token costs through smart routing.
- Kungfu — Maintains work state and context when switching between different AI agents.
- GenericAgent — Grows skills automatically from tasks using a minimal code base.
- MetaClaw — Continuously improves your personal agent from live chats without requiring a local GPU cluster.
All OmniAgent alternatives