Build integrated IS/BS/CF financial workbooks in Excel.
中文简介正在准备中。当前显示原文说明。
openamer/openamer72026年10月19日 更新
免费在 GitHub 公开
mini-openamer
Use when the agent needs recursive reasoning, episodic memory retrieval, internet learning, or world-model prediction. Mini-OpenAmer: the 2B hybrid Mamba core with 9 tools, running 24/7 on Damir's laptop.
中文简介正在准备中。当前显示原文说明。
查看安装方法包含的文件（1）
- SKILL.md2.6 KB
SKILL.md（原文）
安装前可以查看交给 AI 代理的指令内容。
Mini-OpenAmer Integration
Mini-OpenAmer is the self-learning core that runs alongside the main OpenAmer agent. It provides capabilities the main agent can call:
What it provides
- Recursive Reasoning: think → critique → improve loop
- Episodic Memory: 3.012+ episodes with 768d embeddings
- Internet Learning: 5 source types, 24/7 (news, papers, GitHub, docs, competitors)
- World-Model: cause→effect graph with future predictions
- Meta-Learning: adapts its own learning rate and strategy
- Self-Improvement: modifies its own code with test gates
- Swarm Intelligence: laptop + PC as collective (task routing, consensus)
- Auto-Skill Creation: internet insights become new Darwin skills
Integration points
| Component | Location | Purpose |
|---|---|---|
| Tool Server | :8081 | 9 tools + test-time training |
| Smart Router | smart_router.py | 3-tier routing (local→cloud→GPU) |
| Frontier Server | PC :8082 | Qwen3.5-4B deep reasoning |
| Darwin | darwin_engine.py | Skill evolution, 15 min cycles |
| Self-Model | memory/self_model/identity.md | Evolving identity |
| World-Model | memory/world_model.jsonl | Cause-effect + predictions |
| Meta-State | training/meta_state.json | Learning-process self-knowledge |
Energy
- Laptop: ~25 W (2B + all learning loops)
- PC (GPU worker): ~35-50 W (4B warm, training on demand)
- Total: ~60-75 W → target 20 W via quantization/distillation/Akida
- Running cost: 0 € (no API fees, local hardware)
评价
还没有评价。欢迎分享你的使用感受。
评论
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
同一仓库的技能
Run and grow the OpenAmer Agent-to-Agent swarm: identity, trust, node-to-node ask, signed skill/insight sharing, and the autonomous self-learning loop.
中文简介正在准备中。当前显示原文说明。
openamer/openamer72026年10月19日 更新
Use for A/B experiments on OpenAmer configs and skills.
中文简介正在准备中。当前显示原文说明。
openamer/openamer72026年10月19日 更新
Roleplay a hostile user to find and triage UX pain points.
中文简介正在准备中。当前显示原文说明。
openamer/openamer72026年10月19日 更新
Agent mesh: master/worker nodes, HTTP delegation, heartbeat.
中文简介正在准备中。当前显示原文说明。
openamer/openamer72026年10月19日 更新
Use for agentic commerce payments tasks. Candidate synthesized from trend radar; awaiting trial promotion.
中文简介正在准备中。当前显示原文说明。
openamer/openamer72026年10月19日 更新