Federal Award Intelligence — USAspending.gov

Search U.S. federal contracts, grants, loans, IDVs, and direct payments from the official USAspending.gov API and get back clean, normalized, agent-ready JSON — no API key, no anti-bot, public-domain data.

Built for GovCon lead generation, vendor due diligence / KYC, competitive intelligence, and grant prospecting — and designed to be called directly by an AI agent over MCP.

---

Quick start (works with zero configuration)

Click Run with the defaults and you'll get the 100 largest federal contracts awarded in the last 90 days, sorted by dollar value. From there, narrow with:

• Award type — contracts / grants / loans / IDVs / direct payments / other
• Keywords — company name, technology, NAICS description, etc.
• Recipient name contains — find every award to a specific vendor
• Minimum award amount — focus on awards above a dollar threshold
• Look back (days) or explicit Start/End dates
• Place of performance state — e.g. VA

Modes (this is the differentiator)

• Search — normalized award records. The commodity baseline.
• Recipient profiles (vendor intelligence) — rolls awards up by recipient: total won in the window, award count, agencies, and biggest award. This turns raw rows into a decision ("who are the top vendors winning DoD cyber work this quarter") — what a GovCon/BD or due-diligence buyer actually pays for.
• Monitor new awards — on a schedule, returns only awards not seen before (deduped in a persistent store). Wire it to Slack/email for a new-award alert on a competitor or NAICS. Recurring value, not a one-shot dump.

Output (per record)

{
  "award_id": "...",
  "recipient_name": "...",
  "award_amount": 1234567.0,
  "award_type": "DEFINITIVE CONTRACT",
  "contract_award_type": "...",
  "description": "...",
  "start_date": "2026-04-01",
  "end_date": "2027-03-31",
  "awarding_agency": "...",
  "awarding_sub_agency": "...",
  "funding_agency": "...",
  "usaspending_url": "https://www.usaspending.gov/award/<id>/",
  "recipient_id": "...",
  "source": "usaspending.gov",
  "retrieved_at": "2026-06-20T00:00:00.000Z"
}

Use as an MCP tool (agent-native)

Once published, Apify exposes this actor as an MCP tool automatically. Point any MCP client at:

https://mcp.apify.com?tools=<your-username>/federal-award-intelligence

The input fields above become the tool's typed parameters, so an agent can call it autonomously ("find recent DoD cybersecurity contracts over $5M") and get structured records back.

Recommended pricing

Use pay-per-result (PAY_PER_EVENT) at $0.003–$0.005 per award record. Configure the price in the Apify Console under the actor's Publication → Monetization settings, then add a charge call in src/main.js after each pushData: await Actor.charge({ eventName: 'award-record' });. Do not ship it free — on the Store, free signals "hobby project."

Store listing fields (fill these in — they drive discovery)

• seoTitle: USAspending Federal Contract & Grant Search API for AI Agents
• seoDescription: Search US federal contracts, grants, loans & IDVs from USAspending.gov. Clean JSON for GovCon leads, vendor due diligence & competitive intel. No API key.

---

Deploy

npm install -g apify-cli
apify login
# from this folder:
apify push

Then run from the Apify Console, the API, or on a schedule (e.g. daily monitoring for new awards to a competitor).

Honest positioning note

USAspending already has at least one solid incumbent actor on the Store, so treat this as a scaffold and a quality/agent-native play, not an empty field. Your edge is a cleaner normalized schema, MCP-first design, reliable zero-config defaults, and tight GovCon positioning. The same code structure (open API in → normalized records out) transfers 1:1 to a less-crowded data source if your density audit finds a better gap — swap the endpoint, the field map, and the normalization, and keep everything else.

Data: USAspending.gov is public-domain U.S. Government work.