MCP.so
ログイン

Safer Agentic Ai Framework

@NellInc

リポジトリ

Safer Agentic Ai Framework について

概要はまだありません

#ai-safety#agentic-ai#governance

基本情報

カテゴリ

推論

トランスポート

stdio

公開者

NellInc

投稿者

Nell Watson

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "saferagenticai": {
      "command": "uvx",
      "args": [
        "saferagenticai-mcp"
      ]
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Safer Agentic Ai Framework?

Safer Agentic Ai Framework integrates the Safer Agentic AI safety framework into coding assistants via the Model Context Protocol (MCP). It provides 238 normative implementation patterns and 14 operational heuristics across 16 safety suites (9 Drivers + 7 Inhibitors) through 12 read-only tools, enabling developers to search patterns, resolve requirements, map tasks to safety guidance, and follow cross-references without external calls.

How to use Safer Agentic Ai Framework?

Install and run the server using uvx saferagenticai-mcp or pipx install saferagenticai-mcp, then configure your MCP client with the provided command and args. The server exposes 12 read-only tools – such as pattern search, requirement resolution, task-to-safety mapping, and cross-reference following – all accessed via stdio transport with no authentication.

Key features of Safer Agentic Ai Framework

  • Bundles 238 normative patterns and 14 operational heuristics
  • Covers all 16 safety suites (9 Drivers + 7 Inhibitors)
  • 12 read-only tools for search, resolve, map, and cross‑reference
  • No authentication or network calls required
  • Framework data ships with the Python package
  • MIT licensed, requires Python ≥ 3.10

Use cases of Safer Agentic Ai Framework

  • Ensure coding assistant suggestions comply with known safety patterns
  • Resolve ambiguous safety requirements against the normative framework
  • Map a development task to relevant drivers and inhibitors
  • Follow cross-references between patterns and heuristics during code review
  • Quickly check agentic behavior against operational heuristics without leaving the editor

FAQ from Safer Agentic Ai Framework

Does the server require an internet connection?

No. All framework data is bundled in the package – no external calls are made.

What are the runtime dependencies?

Python version 3.10 or later is required. The server uses stdio transport and is MIT licensed.

Is any authentication or API key needed?

No authentication is required. The server is read-only and operates without any credentials.

Can I modify the safety patterns or add new ones?

No. The patterns and heuristics are bundled as read-only data; the server’s 12 tools only provide search, resolution, mapping, and cross-referencing.

Which transport protocols are supported?

Only stdio transport is currently supported.

コメント

「推論」の他のコンテンツ

Agentic Tools MCP Server

Pimzino

A comprehensive Model Context Protocol (MCP) server providing AI assistants with powerful task management and agent memories capabilities with project-specific storage.

87

n8n Workflow Builder MCP Server

salacoste

AI-powered n8n workflow automation through natural language. MCP server enabling Claude AI & Cursor IDE to create, manage, and monitor workflows via Model Context Protocol. Multi-instance support, 17 tools, comprehensive docs. Build workflows conversationally without manual JSON

229

🐢🚀 Node.js Sandbox MCP Server

alfonsograziano

A Node.js–based Model Context Protocol server that spins up disposable Docker containers to execute arbitrary JavaScript.

154

MCP Servers

junfanz1

This project demonstrates a decoupled real-time agent architecture that connects LangGraph agents to remote tools served by custom MCP (Modular Command Protocol) servers. The architecture enables a flexible and scalable multi-agent system where each tool can be hosted independent

27

Agenticstore — The Open Source Standard For Local Mcp Tooling

agenticstore

AgenticStore: The secure toolkit for AI agents. Instantly equip Claude Desktop, Cursor, and Windsurf with 27+ MCP tools, persistent memory, and SearXNG search, all protected by a built-in PII prompt firewall to protect your data from being exposed to AI agents.

4