[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-817":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":30,"readmeContent":31,"aiSummary":32,"trendingCount":16,"starSnapshotCount":16,"syncStatus":33,"lastSyncTime":34,"discoverSource":35},817,"OmniAgent","YeQing17-2026\u002FOmniAgent","YeQing17-2026","An agent capable of self-evolving and dynamically hardening security","https:\u002F\u002Fyeqing17-2026.github.io\u002FOmniAgent\u002F",null,"Python",1767,277,204,25,0,71,189,791,213,20.33,"Other",false,"main",true,[27,28,29],"agent","hermes","open-claw","2026-06-12 02:00:19","\u003Cdiv align=\"center\">\n\u003Cp align=\"center\">\n  \u003Cimg src=\"assets\u002Fomniagent-logo.png\" alt=\"OmniAgent\" width=\"200\">\n\u003C\u002Fp>\n\n# OmniAgent\nAn agent capable of omni-self-evolving and dynamically hardening security\n\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Fyeqing17-2026.github.io\u002FOmniAgent\u002F\">Website\u003C\u002Fa>  •  \n  \u003Ca href=\"https:\u002F\u002Fdocs.omniagent.dev\">Docs (on the way)\u003C\u002Fa>  •  \n  \u003Ca href=\"README.md\">English\u003C\u002Fa>  •  \n  \u003Ca href=\"README_CN.md\">中文\u003C\u002Fa> \n\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPython-3.11+-blue?style=for-the-badge&logo=python&logoColor=white\" alt=\"Python\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-GPL--3.0-blue?style=for-the-badge\" alt=\"License\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FPRs-welcome-brightgreen.svg?style=for-the-badge\" alt=\"PRs Welcome\">\n\u003C\u002Fp>\n\n\u003C\u002Fdiv>\n\n**OmniAgent** is an open-source self-evolving Agent framework inspired by OpenClaw. It's the only agent that implements full-dimensional self-evolution (**OmniEvolve**):\n- **Proactive Memory**: A dual-path alignment mechanism based on explicit interactive feedback and implicit LLM induction enables proactive memory and self-evolving\n- **Skill Self-Evolution**: Through automatic creation, inspection, and repair of skills during interaction, skills evolve in real-time\n- **Context Self-Evolution**: Built on a multi-layer information stack architecture, leveraging real-time user interaction feedback and LLM summarization feedback to continuously update memory and user preferences — achieving self-evolving Personalization Context\n- **BrainModel Self-Evolution**: Through a novel online reinforcement learning feedback loop, the BrainModel iterates dynamically during interaction\n\nTogether, these enable full-dimensional (Skill, Context, BrainModel) self-evolution of the Agent. Additionally, **Hyper Harness** and **Deep Reflexion** modules enhance system safety and task success rate:\n- **Hyper-Harness**: An efficient, safe, and intelligent execution scaffold that provides systematic support for complex tasks\n- **Deep Reflexion**: A dual-layer reflective architecture — real-time risk interception and failure-to-insight conversion — providing a robust guarantee for task success rate\n\n---\n**OmniAgent** V.S. **OpenClaw** V.S. **Hermes**\n\n| Dimension | OpenClaw | Hermes | * OmniAgent |\n| :--- | :--- | :--- | :--- |\n| **Skill Evolution** | Static skills, no evolution | **Periodic** post-execution evolution (slow to take effect) | **Real-time** self-evolution during execution (fast to take effect) |\n| **Skill Injection** | User Message | User Message | User Message (saves 90% token cost) |\n| **Context Evolution** | Static context assembly, no evolution (weak) | Prompt-instruction-based evolution (weak) | Real-time interaction feedback + LLM summarization self-evolution (strong) |\n| **BrainModel Evolution** | Fixed model, no evolution | Fixed model, no evolution | Self-deployed model, online RL evolution |\n| **Harness Safety** | Static security scanning (bypassable) | Skill trust-level policy, static scanning (bypassable) | **Tool & Skill** trust-level policy + four-layer dynamic security scanning (unbypassable) |\n| **Hyper-Harness** | None (slow) | None (slow) | Dynamic multi-agent + dynamic concurrent tool execution (fast) |\n| **Agent-Loop** | ReAct single loop (low success rate) | ReAct single loop (low success rate) | Dual-layer Deep Reflexion loop (high success rate) |\n\n\n## Core Features\n\n**OmniEvolve (Full-Dimensional Self-Evolution)**: The agent evolves continuously through interaction, and safety hardens dynamically.\n\n- **Proactive Memory**: Based on a multi-layer information stack, a dual-path alignment mechanism of explicit user feedback and implicit LLM induction enables autonomous precipitation and continuous self-evolution of user profiles and memory\n- **Skill Self-Evolution**: Through pattern extraction from high-frequency action sequences, skills are natively auto-generated; leveraging dual-path feedback from user interaction and LLM diagnosis, skills are automatically diagnosed and repaired\n- **Personalization Context**: Through real-time capture of multi-dimensional preference signals, an adaptive personalized context is constructed, achieving precise alignment between the Agent-Loop and individual user preferences\n- **BrainModel Self-Evolution**: Through a novel online reinforcement learning (GRPO + PRM) feedback loop, the BrainModel achieves closed-loop self-evolution during interactive use\n\n**Hyper-Harness (Super Scaffold)**: A more efficient, safe, and intelligent Harness engine.\n\n- **Progressive Context Loading**: A design pattern inspired by Anthropic Claude Skills — Progressive Disclosure — loading on demand in graduated stages\n- **Dynamic Multi-Agent**: Introducing **Sentinel** (planning) and **Guardian** (safety) agents that dynamically analyze task complexity and risk level, activating in real-time to improve success rate and safety\n- **Dynamic Concurrent Tool Execution**: Auto-resolves inter-tool dependencies, shifting from serial waiting to async parallel invocation, reducing latency for long-chain tasks\n- **Four-Layer Dynamic Security Scanning**: LLM intelligent review → Policy engine → Interactive approval → Execution sandbox. Through trust-level classification, different Skills apply different security policies. Security scanning is unbypassable (industry-first)\n\n**Deep Reflexion (Inner-Outer Dual-Layer Reflective Architecture)**: Improves agent task success rate (PASS@1).\n\n- **Inner-to-Outer Failure Experience Conversion**: Based on LLM-driven automatic root cause analysis (RCA) and heuristic strategy extraction, Reflexion is dynamically injected into the context space, achieving an inner-outer dual-layer collaborative closed-loop reflective correction and intelligent retry\n- **Inner Failure Prevention Mechanism**: Three-layer failure prevention system (trajectory repetition, error action repetition, loop pseudo-termination) monitors failure risks and injects context, improving task success rate (PASS@1)\n\n---\n\n## What Can You Do With OmniAgent\n\n| Use Case | What OmniAgent Does |\n|----------|-------------------|\n| **Workspace & Skills** | Config injection: define Agent personality, tasks, and behavior rules via bootstrap files (AGENTS.md \u002F SOUL.md \u002F CUSTOM.md); Progressive loading: read associated documents in graduated stages (L0\u002FL1\u002FL2) based on conversation depth to prevent Token overflow |\n| **Coding & Dev** | Full-lifecycle code handling: write, run, and test code directly in the local environment; Auto-correction: on runtime errors, the Agent reads Traceback and attempts fixes until the program runs |\n| **Research & Analysis** | Multi-source web search: auto-invokes search tools and visits multiple pages to extract key information; Knowledge cross-validation: compares information from different sources, outputs a comprehensive report with source annotations |\n| **System Admin** | Shell command execution: supports terminal commands in sandbox or host environments; Safety control flow: built-in security scanning system auto-suspends and requests user approval for high-risk commands like delete and format |\n| **Multi-Channel** | Unified gateway: manages message routing for Feishu, Discord, Telegram, CLI, and more; Session persistence: seamless switching between clients while maintaining Agent memory consistency |\n| **Flexible LLM Backends** | Hybrid model routing: freely combine OpenAI, Claude, DeepSeek, Ollama, and other backends |\n\n---\n\n## Quick Start\n\n**Requirements:** Python 3.11+, an LLM API key (DeepSeek \u002F OpenAI \u002F Anthropic \u002F Ollama \u002F Gemini).\n\n### Installation\n\n```bash\n# 1. Install\n$ pip install -e .\n\n# 2. Interactive setup — choose provider, enter API key, done\n$ omniagent onboard\n\n# 3. Start\n$ omniagent chat                    # CLI\n$ omniagent serve                   # Web UI → http:\u002F\u002F127.0.0.1:18790\n```\n\n### Three Ways to Interact\n\n| Mode | Command | Description |\n|------|---------|-------------|\n| **Terminal** | `omniagent chat` | Interactive chat in your terminal |\n| **Web UI** | `omniagent serve` | Start Gateway, open http:\u002F\u002F127.0.0.1:18790 in your browser |\n| **Mobile** (Feishu \u002F Discord \u002F Telegram) | `omniagent serve` | Start Gateway, configure Channel in `config.yaml`, then open a session in your terminal |\n\n---\n\n## Configuration\n\nConfiguration is layered: defaults → `~\u002F.omniagent\u002Fconfig.yaml` → environment variables.\n\n```yaml\nproviders:\n  deepseek:\n    api_key: \"sk-your-key\"\n    model_id: deepseek-chat\n\nagent:\n  model_provider: deepseek\n  reflexion_enabled: true\n```\n\n> Full configuration reference: [docs.omniagent.dev](https:\u002F\u002Fdocs.omniagent.dev) *(coming soon)*\n\n---\n\n## Architecture & Project Structure\n\n```\n┌──────────────────────────────────────────────────────┐\n│  Channels:  CLI · Web UI · Feishu · Discord · Telegram│\n└──────────────────────┬───────────────────────────────┘\n                       │\n┌──────────────────────▼───────────────────────────────┐\n│  Gateway  (WebSocket + HTTP · Session Management)     │\n└──────────────────────┬───────────────────────────────┘\n                       │\n┌──────────────────────▼───────────────────────────────┐\n│  Reflexion Agent-Loop                                │\n│                                                      │\n│  ┌─────────────────┐  ┌──────────────────────────┐   │\n│  │ Deep Reflexion  │  │ Hyper-Harness            │   │\n│  │ Reflexion Loop  │  │ ┌────────────────────┐   │   │\n│  │ Failure Prevent │  │ │Progressive Loading │   │   │\n│  └─────────────────┘  │ │Dynamic Tool Exec   │   │   │\n│                       │ │4-Layer Sec Scan    │   │   │\n│  ┌─────────────────┐  │ │Dynamic Multi-Agent │   │   │\n│  │ Sentinel Agent  │  │ └────────────────────┘   │   │\n│  │ (Planning)      │  └──────────────────────────┘   │\n│  ├─────────────────┤                                 │\n│  │ Guardian Agent  │  ┌──────────────────────────┐   │\n│  │ (Safety Review) │  │ OmniEvolve               │   │\n│  └─────────────────┘  │ ┌────────────────────┐   │   │\n│                       │ │Proactive Memory    │   │   │\n│                       │ │Skill Self-Evolution│   │   │\n│                       │ │Personalization     │   │   │\n│                       │ │BrainModel Self-Evo │   │   │\n│                       │ └────────────────────┘   │   │\n│                       └──────────────────────────┘   │\n└──────────────────────────────────────────────────────┘\n                       │\n┌──────────────────────▼───────────────────────────────┐\n│  LLM Providers                                       │\n│  DeepSeek · OpenAI · Anthropic · Ollama · Gemini     │\n│  OpenRouter · vLLM · SGLang · Custom                 │\n└──────────────────────────────────────────────────────┘\n```\n\n```\nomniagent\u002F\n├── agents\u002F          # Core: reflexion loop, sentinel, guardian, skill\u002Fmemory evolution, context management\n├── security\u002F        # Policy engine, approval, audit, sandbox\n├── tools\u002F           # Built-in tools\n├── channels\u002F        # Feishu, Discord, Telegram, Webhook\n├── config\u002F          # OmniAgentConfig + sub-configs\n├── gateway\u002F         # WebSocket + HTTP server\n└── rl\u002F              # GRPO + PRM training pipeline\n```\n\n---\n## Roadmap\n\n### Near-term\n\n- [ ] **Four-Layer Proactive Memory System (Proactive Memory 2.0)** — Automatically extract and persist long-term memories from conversations, refine the proactive memory system, and resolve conflicts between self-evolution rules\n- [ ] **Agent Plan-Mode** — Implement a new Agent planning mode that generates and confirms a plan before executing complex tasks, then executes step by step\n- [ ] **Channel Ecosystem Expansion** — Add WeChat, WeCom, DingTalk, and more connectors; improve channel abstraction layer to reduce integration cost\n- [ ] **Multi-Agent Collaboration** — Enhance inter-agent communication and task orchestration, support dynamic delegation and result aggregation between agents\n- [ ] **Documentation**\n- [ ] **More Tests**\n\n---\n## Contributing\n\n1. Fork the repo and create a feature branch\n2. Make your changes with tests\n3. Run `python -m pytest` to verify\n4. Submit a pull request\n\nWe welcome contributions of all kinds — bug fixes, new tools, channel connectors, and documentation improvements.\n\n---\n## License\n\nThis project is licensed under [GPL-3.0](https:\u002F\u002Fwww.gnu.org\u002Flicenses\u002Fgpl-3.0.txt). Any code that references this project must also be open-sourced under the same license.\n\n\n\n---\n## Star History\n\n\u003Ca href=\"https:\u002F\u002Fwww.star-history.com\u002F?repos=YeQing17-2026%2FOmniAgent&type=date&legend=top-left\">\n \u003Cpicture>\n   \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fchart?repos=YeQing17-2026\u002FOmniAgent&type=date&theme=dark&legend=top-left\" \u002F>\n   \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fchart?repos=YeQing17-2026\u002FOmniAgent&type=date&legend=top-left\" \u002F>\n   \u003Cimg alt=\"Star History Chart\" src=\"https:\u002F\u002Fapi.star-history.com\u002Fchart?repos=YeQing17-2026\u002FOmniAgent&type=date&legend=top-left\" \u002F>\n \u003C\u002Fpicture>\n\u003C\u002Fa>\n\n---\n\u003Cdiv align=\"center\">\n\n**OmniAgent** — An agent whose intelligence evolves with every interaction, and whose safety hardens dynamically.\n\n\u003C\u002Fdiv>\n","OmniAgent 是一个能够全方位自我进化并动态增强安全性的代理框架。它基于Python开发，核心功能包括通过主动记忆、技能自我进化、上下文自我进化以及脑模型自我进化实现全维度的自我提升。特别地，OmniAgent 采用了双路径对齐机制来促进主动记忆，并通过在线强化学习反馈循环使脑模型能够动态迭代。此外，Hyper Harness 和 Deep Reflexion 模块进一步增强了系统的安全性和任务成功率。该项目适用于需要高度自适应性与安全性保障的应用场景，如复杂任务执行环境下的自动化助手或智能代理服务。",2,"2026-06-11 02:39:33","CREATED_QUERY"]