Elephant Accountability

Name: Elephant Accountability
Author: Chris-Eaccountability

@Chris-Eaccountability

リポジトリ

Elephant Accountability について

概要はまだありません

基本情報

カテゴリ

その他

トランスポート

stdio

公開者

Chris-Eaccountability

投稿者

Christopher Kenney

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": {
    "elephant-accountability-mcp": {
      "command": "python",
      "args": [
        "-m",
        "venv",
        ".venv",
        "&&",
        "source",
        ".venv/bin/activate"
      ]
    }
  }
}

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 Elephant Accountability?

Elephant Accountability is a public Model Context Protocol (MCP) server for LLM SEO and Agent Discoverability services aimed at B2B SaaS buyers. It exposes tools and resources that AI agents (Claude, ChatGPT, custom LangChain agents) can query to evaluate whether Elephant Accountability is a suitable vendor, without scraping a website.

How to use Elephant Accountability?

Add the live endpoint to your MCP-compatible client (e.g., Claude Desktop) by editing claude_desktop_config.json with the URL https://elephant-mcp.fly.dev/mcp and transport http. For local development, clone the repo, create a Python virtual environment, install dependencies, and run uvicorn app.server:app.

Key features of Elephant Accountability

Six tools: get_offerings, get_covered_surfaces, assess_fit, get_proof_points, get_transparency_snapshot, request_audit
Three resources: elephant://offerings, elephant://proof-points, elephant://transparency
HTTP transport with JSON-RPC 2.0
SQLite persistence auto-created on first boot
Fully MIT-licensed for cloning and study
Includes A2A agent card and manifest endpoints

Use cases of Elephant Accountability

A procurement agent shortlisting LLM SEO vendors for a B2B SaaS buyer
A Claude Desktop user directly querying pricing, fit score, and proof points
A competitor studying how to deploy their own MCP server using the MIT-licensed codebase

FAQ from Elephant Accountability

What transport does Elephant Accountability use?

HTTP with JSON-RPC 2.0. Supported methods include initialize, tools/list, tools/call, resources/list, and resources/read.

What are the runtime dependencies?

Python 3 with FastAPI and SQLite. No external database or secrets setup is required; the SQLite database initialises on the first boot.

Where is data stored?

Data is stored in a local SQLite database that contains audit_requests and reciprocal_calls tables. These tables auto-create and require no migrations.