bidsparq-mcp-server
Query federal RFP, subaward, pricing, and vendor data from AI clients via 55 tools, including procurement contact graph and contract vehicle intelligence.
README
bidsparq-mcp-server
Query BidSparq's federal RFP, subaward, pricing, and vendor data from Claude Desktop, Claude Code, Cursor, Cline, and any other MCP-compatible AI client.
Pro Max plan required. Get one at bidsparq.com/plan.
What you get
- Find primes who hire subcontractors in your NAICS (102K+ FFATA subaward records, 18-month window)
- Pricing percentiles (p25 / median / p75) + agency average + year-over-year trend
- Wired-RFP detection (brand-name lock, intent-to-award, thin PWS, competitor-specific language)
- Hybrid RFP search (keyword + semantic + full-text PDF) across 130+ sources
- Past contract awards, recompete tracking, agency profiles, seasonal patterns
- SAM-registered vendor lookup (NAICS, set-asides, certifications, UEI, location)
- Buyer-side procurement-officer contact graph (Beacon-equivalent): find contacts at a given agency, deep-dive on any contact's posting history (85,978 procurement officers from real solicitation activity)
- Federal contract vehicle intelligence: 60,000+ active GWACs, GSA Schedules / FSS, BPAs, IDCs, BOAs — search by NAICS, agency, set-aside, vehicle type; get personalized recommendations; track expiration windows as recompete signals
All 55 tools available from your AI client. Unlimited usage on Pro Max subject to fair-use limits (per-minute / per-hour / daily cost), shared with in-app chat.
Install
1. Get an API key
bidsparq.com/settings/api-keys → Create Key → copy the bsq_live_... value.
2. Add to your MCP client
Claude Desktop
Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%/Claude/claude_desktop_config.json (Windows):
{
"mcpServers": {
"bidsparq": {
"command": "npx",
"args": ["-y", "bidsparq-mcp-server"],
"env": {
"BIDSPARQ_API_KEY": "bsq_live_..."
}
}
}
}
Restart Claude Desktop.
Claude Code (CLI)
Add to .mcp.json in your project (or ~/.claude/mcp.json globally):
{
"mcpServers": {
"bidsparq": {
"command": "npx",
"args": ["-y", "bidsparq-mcp-server"],
"env": {
"BIDSPARQ_API_KEY": "bsq_live_..."
}
}
}
}
Cursor
Edit .cursor/mcp.json in your project:
{
"mcpServers": {
"bidsparq": {
"command": "npx",
"args": ["-y", "bidsparq-mcp-server"],
"env": {
"BIDSPARQ_API_KEY": "bsq_live_..."
}
}
}
}
Cline / RooCode / Continue
Same JSON shape — refer to your extension's MCP config docs and paste the snippet above.
Try it
After install, ask your AI client:
- "Find me primes who hire subs in NAICS 541512"
- "What's the typical award amount for VA software contracts?"
- "Show me wired RFPs in defense IT so I know what to avoid"
- "What contracts are coming up for recompete in cybersecurity?"
- "Tell me about NASA's buying patterns"
- "Find HUBZone businesses in Texas for NAICS 541512"
Configuration
| Env var | Required | Default | Notes |
|---|---|---|---|
BIDSPARQ_API_KEY |
yes | — | bsq_live_... from your settings page |
BIDSPARQ_BASE_URL |
no | https://bidsparq.com |
Override for self-hosted / staging |
Troubleshooting
Invalid BIDSPARQ_API_KEY— Key is wrong, expired, or revoked. Create a new one in your settings.Pro Max plan required— Upgrade at bidsparq.com/plan.Rate limit exceeded— Pro Max has fair-use limits to protect against runaway scripts (per-minute: 60, per-hour: 500, daily cost cap). The 429 response includes aRetry-Afterheader. Most users will never see this; if you do, slow your loop or contact support.BidSparq MCP is not yet enabled in your account— MCP is in staged rollout. Email hello@bidsparq.com for early access.
Privacy
- The MCP server only sends data you explicitly query for (NAICS, agency names, search terms).
- Your conversations stay in your AI client — BidSparq sees only the tool calls.
- API keys are bcrypt-hashed at rest; only the first 12 chars (
bsq_live_xxx) are visible after creation.
License
MIT. Source: https://github.com/n0edlg/bidsparq-mcp-server
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。