Government Procurement Feeds
LocalAbout Government Procurement Feeds
Government Procurement Feeds is an MCP server for searching Mississippi and UK government tenders and tracking changes. It serves live procurement data through a shared set of tools, including a feed of deadline updates, awards, and added documents.
Key features
- 01Search Mississippi and UK opportunities
- 02Retrieve full opportunity records and document links
- 03Track deadline changes, awards, and added documents
- 04Choose a feed with SLED_SOURCE
- 05Run as a Python standard-library server
Use cases
- 01Find open government bids by keyword, category, or date
- 02Review tender records and official notices
- 03Monitor changes to deadlines, awards, and documents
MCP support
The server exposes search_opportunities, get_opportunity, and list_recent_changes to MCP clients over stdio. The feeds are also available through Apify's hosted MCP server.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
We haven’t collected a package or remote endpoint for this server. Follow the instructions in the project README.
Tools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMEsled-feed — Government procurement feeds with change tracking (MCP)
Search live US (Mississippi) and UK (Contracts Finder) government bids, RFPs and tenders — and get the one thing the official portals won't push you: what CHANGED. Deadline extensions, new awards, and posted addenda, exposed as Model Context Protocol (MCP) tools so an AI agent can query procurement data directly.
- 🇺🇸 Mississippi state government solicitations (official DFA Contract Bid Search)
- 🇬🇧 UK Contracts Finder tenders (official Cabinet Office OCDS API, Open Government Licence v3.0)
Both feeds expose the same three MCP tools. The standalone server (mcp_server.py) is a
thin client that runs the live Apify Actors on demand, so a fresh
clone returns real data with no local database to seed.
Tools
| Tool | What it does |
|---|---|
search_opportunities | Filter live tenders/bids by category, keyword, status, posted/due dates. |
get_opportunity | Full record by id, including all document links + the official notice URL. |
list_recent_changes | The differentiator — deadline changes, new awards (Open → Awarded), added documents over time. |
Quickstart (standalone MCP server, stdio)
The server is pure Python standard library — no pip install required. You only need
an Apify API token (free tier works); each call runs the live Actor on your own account.
git clone https://github.com/tobyrichmond12-max/sled-feed.git
cd sled-feed
export APIFY_TOKEN=<your Apify API token> # required — get one at apify.com (free)
export SLED_SOURCE=mississippi # or: uk_contracts_finder
python3 mcp_server.py # speaks MCP over stdio (JSON-RPC 2.0)
# sanity check without a client:
python3 mcp_server.py --selftest
MCP client config (Claude Desktop / VS Code / Cursor)
{
"mcpServers": {
"sled-feed-mississippi": {
"command": "python3",
"args": ["/absolute/path/to/sled-feed/mcp_server.py"],
"env": { "APIFY_TOKEN": "YOUR_APIFY_TOKEN", "SLED_SOURCE": "mississippi" }
},
"sled-feed-uk": {
"command": "python3",
"args": ["/absolute/path/to/sled-feed/mcp_server.py"],
"env": { "APIFY_TOKEN": "YOUR_APIFY_TOKEN", "SLED_SOURCE": "uk_contracts_finder" }
}
}
}
No-install alternative: Apify hosted MCP
Both feeds are also callable through Apify's hosted MCP server — no clone, no local process:
{
"mcpServers": {
"apify-gov-feeds": {
"url": "https://mcp.apify.com/?tools=toby_richmond/mississippi-procurement-feed,toby_richmond/uk-contracts-finder-feed",
"headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
}
}
}
Example usage
// search open IT bids
{ "name": "search_opportunities",
"arguments": { "status": "Open", "keyword": "information", "limit": 5 } }
// the change feed (the differentiator)
{ "name": "list_recent_changes",
"arguments": { "since": "2026-06-01T00:00:00" } }
Example record (Mississippi):
{
"solicitation_id": "1601-26-R-RFIN-00066",
"title": "SPIRIT Management Information System ...",
"agency": "Mississippi Department of Health",
"category": "INFORMATION TECHNOLOGY (IT)",
"status": "Open",
"due_date": "2026-06-30T15:00:00-05:00",
"due_timezone": "America/Chicago",
"document_links": [{ "label": "Solicitation PDF", "url": "https://.../FORM.PDF" }],
"source_url": "https://www.ms.gov/dfa/contract_bid_search/Bid/Details/45668"
}
Example change event (list_recent_changes):
{ "event": "deadline_changed", "solicitation_id": "1601-26-R-RFIN-00066",
"old_due_date": "2026-06-30T15:00:00-05:00", "new_due_date": "2026-07-07T15:00:00-05:00" }
Data, schema & privacy
Per solicitation: solicitation_id, title, agency (buying organisation only),
category, bid_type, posted_date / due_date (ISO-8601 with explicit timezone and
closing time), status (Open / Closed / Awarded / Archived), document_links,
source / source_url. The UK feed also carries estimated_value + value_currency.
No personal data. Buyer contact names/emails/phones in the source are stripped at ingestion and re-checked at storage; only solicitation data and the buying organisation name are served. A built-in monitor watches for upstream breakage (renamed fields, response-shape changes, record-count crashes, new status values, or PII drifting toward output) and returns "temporarily unavailable" rather than serving stale or wrong data.
How it works
MCP client ──stdio──> mcp_server.py ──HTTPS──> Apify Actor (run-sync) ──> official gov API
(thin client) (ingest → diff → change-log → serve)
The Actors persist a SQLite snapshot in a named Apify key-value store between runs, so the
change history (list_recent_changes) accumulates run over run. Source code for the
pipeline, monitor, and storage layer is all in this repo.
License
Code: MIT (see LICENSE). UK Contracts Finder data is provided under the Open Government Licence v3.0 (commercial reuse permitted). Mississippi data is public-record government information.
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.