tenderapi-mcp
Exposes French and EU public procurement data (BOAMP + TED) as MCP tools for AI agents, enabling search for tenders, awards, and winner intelligence via typed filters.
README
TenderAPI MCP server
<!-- mcp-name: io.github.IDNSIDNS/tenderapi-mcp -->
Expose TenderAPI (French BOAMP + EU TED public procurement data) as MCP tools for AI agents — Claude Desktop, Cursor, Continue, Zed, etc.
A thin wrapper over the public REST API at https://tenderapi.fr.
Install
Requires Python 3.10+.
From PyPI (once published):
pip install tenderapi-mcp
From source:
git clone https://github.com/IDNSIDNS/tenderapi-mcp
cd tenderapi-mcp
pip install -e .
Configure
Get a free API key at https://tenderapi.fr/.
Set the env var:
export TENDERAPI_KEY=ta_your_key_here
Use with Claude Desktop
Edit your Claude Desktop config:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json - Linux:
~/.config/Claude/claude_desktop_config.json
{
"mcpServers": {
"tenderapi": {
"command": "tenderapi-mcp",
"env": {
"TENDERAPI_KEY": "ta_your_key_here"
}
}
}
}
Restart Claude Desktop. The tenderapi server should appear in the tool picker.
Use with other MCP clients
Any MCP client supporting stdio transport. The binary tenderapi-mcp (installed by pip) is the entry point.
Tools exposed
| Tool | Tier | Description |
|---|---|---|
search_tenders |
Free | Search BOAMP + TED tenders with typed filters (CPV, region, budget, deadline, source…) |
search_awards |
Starter | Search award notices (who won which contract, for how much) |
winner_intel |
Pro | Aggregated winner stats — top companies by CPV / region / year |
me |
— | Current key tier, quota remaining, available features |
Tiers
- Free: 100 req/day — tenders only
- Starter (5 €/mo HT): 5 000 req/day — adds awards + webhooks
- Pro (15 €/mo HT): 50 000 req/day — adds winner intelligence
See https://tenderapi.fr/#pricing.
Local development
Override the API base URL via TENDERAPI_BASE_URL (default https://tenderapi.fr).
Changelog
0.1.2 (2026-05-07)
The underlying TenderAPI changed behaviour in ways that affect MCP clients:
- Stricter query parameters. Unknown parameters (e.g.
q=instead ofkeyword=) now return400 Bad Requestwith a list of valid parameters, instead of being silently ignored. Update any tool calls that relied on the old behaviour. - Multi-word
keywordsearch now AND-matches tokens.keyword=espaces vertsrequires both words to appear (in title or description). Previously it required the literal substring "espaces verts". - Comma-separated lists supported on
regionandcpv.region=paca,bretagnenow works as well as repeating the parameter. - TED awards now expose real SIRETs (
winner_siret,buyer_siret) for ~50% of records. BOAMP SIRETs are being progressively enriched via the SIRENE register. - Country codes normalised to ISO-3166-1 alpha-2 (
FR, notFRA) across both BOAMP and TED awards/tenders.
0.1.1 (2026-05-04)
Initial public release on the MCP Registry.
License
MIT — see LICENSE.
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。