mssclaw v0.3.0
世界上第一个内置「意义场自检」的开源 Agent 框架
mssclaw (框架平台)
├── mss-ai models ← 本地模型 (mss-ai-v3.4.3等, 9 models)
├── Agent 引擎 ← L2护城河 + 6流式模式 + 10工具 + RAG
├── Vault 保险箱 ← 密码管理全栈 (AES-256-GCM)
├── Library 库系统 ← 8库643条目统一管理
│ ├── tools(10) + skills(5) + models(30) + kb(618)
│ └── immunity + agents + sessions + custom
├── Multi-Model 编排 ← Shell模式(壳+核) + Pipeline + AutoRoute
└── CLI ← vault|chat|serve|absorb|library|models|status
🛡️ L2 独有护城河
- 🔥 A3 热税预算 — 自动拒绝无意义任务
- Δ 意义开放度检测 — 实时监控闭合/循环
- 🛡️ 规范场+幻觉盾 — 31规则+4类检测
- 🧠 认知框架 — 能力自知+身份锚定+演化就绪
- 📊 Δ 健康监控 — 不只"活着", 要"有意义"
🤖 Agent 核心能力
- LLM 后端: Ollama + OpenAI兼容
- 流式输出: 6模式+语义感知+深度折叠+速度对齐
- 工具调用: 10内置工具+L2安检 (safe|file|office)
- RAG 管道: BM25+密度, 零外部依赖
- 多Agent流水线: Writer→Reviewer→Refiner
- 容错: 重试+降级+熔断
- 记忆: 三层存储+自动凝聚
- 评测: 道评分 valid−pseudo×2.0 (MSS独有)
🔐 MSS Vault 全栈
- 加密: AES-256-GCM, Zero-Trust
- CLI:
mss-vault15+命令 - Web面板: http://127.0.0.1:5099
- HTTP API: RESTful
- 导入: Chrome/Edge 一键迁移
- 备份: 自动+旋转
🌐 全球模型目录 (30 models)
- Cloud(18): GPT-4o, Claude 3.5, Gemini 2.0/2.5, DeepSeek V3/R1, Grok-2...
- Local(9): Ollama自动扫描
- MSS(3): mss-ai-v3.4.3-balanced/slim, mss-ai-v3.4.2-production
🧬 MSS 独有特性
| 特性 | 同行 |
|---|---|
| L2 意义场护城河 | ❌ |
| 语义流式+折叠+路由 | ❌ |
| 道评分评测 | ❌ |
| 技能编译器(吸收→消化) | ❌ |
| 逻辑病毒检测+群体免疫 | ❌ |
| Shell双模型(壳+核) | ❌ |
| 多模型智能编排 | ⚠️ 部分 |
| 统一库管理+跨库检索 | ❌ |
🚀 快速开始
pip install mss-agent
mss-vault quickstart
mssclaw chat --model qwen2.5:7b
mssclaw serve --all
📊 开发统计
- 81 Sprints | 117 Tests | 81 Commits
- v0.3.0 | GitHub: mysama1/MSS-AI-Project
- License: MIT
1 maintainer
Metadata
1 maintainer
Release files for mss-agent 0.3.10
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| mss_agent-0.3.10-py3-none-any.whl | Python 3 | none | any | Details for mss_agent-0.3.10-py3-none-any.whl |
Release files / mss_agent-0.3.10-py3-none-any.whl
| Download URL | mss_agent-0.3.10-py3-none-any.whl |
|---|---|
| Size | 1.1 MB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
896969104af84dc0213f638b922b06f9db3a0520f2ede576719e7c3ab5193fd4
|
|
BLAKE2b-256 checksum How to use checksums |
19d362e28d439de6d2503d68d21bd80aee94c89e79218945c481d16ae92f2bbf
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.11.9
|
Security for mss-agent 0.3.10
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.
