Defense Explorer for ChatGPT
ChatGPT can use remote MCP servers as custom connectors on eligible plans. Add Defense Explorer and ask about defence tenders, fit scores and tender documents with sources attached.
In short
If your ChatGPT plan and workspace allow custom MCP connectors, add the Defense Explorer URL, sign in with OAuth, and ChatGPT can search and read defence tender data under your workspace permissions.
- 01Custom MCP connectors in ChatGPT depend on your plan and, in business workspaces, on admin settings. OpenAI has exposed them through a developer mode setting.
- 02Use OAuth: ChatGPT discovers the Defense Explorer authorization server from the URL.
- 03The server behaves the same in every client, so answers match what Claude or the web app would show.
Connect in five minutes
[ 03 STEPS ]OpenAI renames and moves these settings. We describe the flow, not the exact menu; OpenAI’s help centre has the current path.
- STEP 01
Check that custom connectors are available
Look in ChatGPT’s settings for connectors or apps. Adding your own MCP server may require turning on developer mode, and on a business or enterprise workspace an admin may need to allow custom connectors first.
- STEP 02
Add Defense Explorer as a connector
Create a new custom connector, name it Defense Explorer, paste the server URL and choose OAuth as the authentication method. ChatGPT sends you to a Defense Explorer sign-in, where you can reduce the scopes before approving.
Name: Defense Explorer MCP server URL: https://api.defenseexplorer.com/mcp Authentication: OAuth - STEP 03
Use it in a chat
Enable the connector for the conversation, then ask “Which procurement sources can you see?”. ChatGPT should call list_sources. If it cannot see the tools, check the connector is switched on for that chat.
Tools ChatGPT gets
[ CORE TOOLS ]The core tools. The [full list](/integrations/mcp/) is on the server page.
| Tool | Scope | What it returns |
|---|---|---|
| search_opportunities | mcp:read | Normalized, deduplicated notices filtered by query, source, type, country, CPV, value and dates, with official source links and English titles where translated |
| match_opportunities | mcp:read | Notices ranked against your company profile, each with its 0–100 score, reasons, unknowns and blockers |
| ask_opportunity | mcp:ai | A grounded answer about one notice, with sources and a document coverage note; facts are kept apart from inferences |
| get_brief | mcp:read | The cited brief: each claim with the document span it cites and the verbatim quote, plus a stale flag |
| get_document_coverage | mcp:read | Documents registered for a notice, their parse status and an explicit limitations list |
| list_pursuits / get_pursuit | mcp:read | The notices your team is working on, with stage, owner, next action and deadline |
| list_changes / get_change | mcp:read | Material amendments on followed notices, with original and updated clause text side by side |
| list_market_signals / get_market_signal | mcp:read | Early signals: prior information notices, market consultations, stated contract end dates, programme plans |
| list_sources | mcp:read | Every indexed source: coverage, freshness, counts, open incidents and known limitations |
| record_decision | mcp:write | Records a pursue, hold or decline decision a person has made, with their rationale, to the audit trail |
Write tools may ask you to confirm before ChatGPT runs them, depending on your ChatGPT settings. Grant mcp:write only if you want watchlist saves and decision records from chat.
Things to ask ChatGPT
[ 06 PROMPTS ]Realistic questions from bid and BD teams, and the tool behind each answer.
- 01match_opportunities
“Which open NSPA tenders close in the next 30 days that fit our radio product line?”
Ranks open notices against the product line in your company profile and returns each score with its reasons, unknowns and blockers, so the assistant can say why a notice is on the list.
- 02get_brief · ask_opportunity
“Summarise the brief for this notice and tell me what the buyer wants as proof of past delivery.”
Every claim in the brief carries a page reference and a verbatim quote. Claims marked unknown or conflict come back labelled, not as facts.
- 03get_requirement_assessments
“Can we meet every mandatory requirement on the Dutch secure comms pursuit?”
Returns the qualification matrix. Only approved evidence inside its validity dates satisfies a requirement, and a blocked item names the rule that blocked it.
- 04list_changes · get_change
“What changed on the tenders we are following this week, and which requirements need a second look?”
Material amendments only, with the original and updated clause side by side. Cosmetic edits are excluded by default.
- 05summarize_market_signals
“What demand is building for counter-UAS in Germany and France over the next year?”
Counts prior information notices, market consultations and stated contract end dates by horizon. These are inferred demand, not open calls, and the assistant should say so.
- 06list_sources
“Do you cover Spanish Ministry of Defence tenders, and how fresh is that data?”
States what each source covers and leaves out, when it last succeeded and any open incident. This is the answer to “do you cover this?”, before anyone concludes a notice does not exist.
Getting reliable answers
[ PRACTICE ]Ask for the notice id and the source link with every answer, so anyone can check the official record. For document questions, ask ChatGPT to include the quote from the brief and to say when document coverage is incomplete.
Match scores come from the company profile in Defense Explorer, not from the chat. If a score looks wrong, fix the profile or give relevance feedback in the workspace, and every client improves at once.
Remember that market signals are demand that may become a tender, not open calls. A good answer about signals says which dates were stated by a source and which are estimates.
Safety and permissions
[ WHAT IT WILL NOT DO ]An intelligence and workflow surface, not an autonomous procurement agent.
The server reads public procurement records and your workspace’s own data under the same company membership and permissions as the web app. Another workspace’s id returns “not found”, never “forbidden”, so ids cannot be probed. Access stops as soon as a membership ends or a token is revoked or expires.
Writes are deliberately narrow and all need mcp:write: saving or removing a watchlist item, recording a bid decision a person has already made, and acknowledging that an amendment was seen. An assistant cannot move a pursuit through stages on its own, approve evidence, edit a requirement or create an export.
It cannot submit bids, contact buyers, draft proposal text, access classified material, change billing or make arbitrary network requests. Raw source snapshots are never returned directly, because source pages and documents are untrusted input; ask_opportunity is the controlled path for reading indexed tender text.
Guard rails the tools make visible
A claim with status unknown or conflict is not a supported fact. Document coverage limitations matter: “no such clause” and “no such clause in the parts we could read” are different answers. Only an approved evidence item inside its validity dates can satisfy a requirement. Market signals are inferred demand, not open calls, and an estimated window is labelled as an estimate.
Coverage is defence-relevant, not complete national procurement, so list_sources should be checked before an assistant says a notice does not exist.
Other tender tools with MCP
[ HONEST CONTEXT ]MCP is becoming standard in this category. What differs is the data and the tools behind it.
Defense Explorer is not the only tender tool with an MCP connection. Tenderlake offers an MCP-based analyst for mainstream assistants, and Stotles added an MCP connector on its paid plans in August 2026. If you are comparing, see Defense Explorer vs Tenderlake and Defense Explorer vs Stotles.
What is specific here: notices classified for defence relevance and clearance evidence with the reason shown; NATO/NSPA and national European sources alongside TED; briefs that cite the page and quote the text; a qualification matrix tied to reviewed company evidence; amendment impact; and write tools limited to recording what a person decided.
Questions
[ 05 ANSWERS ]01Does ChatGPT support MCP servers?
Yes, as custom connectors on eligible plans, with availability and menu names set by OpenAI and, in workspaces, by admins. Check OpenAI’s help centre for the current requirements.
02Can I use a dx_mcp_ token instead of OAuth in ChatGPT?
Use OAuth. It is the path ChatGPT’s connector flow is built around, it gives short-lived tokens, and you can reduce scopes when you sign in. Tokens suit clients that take custom headers, such as Cursor.
03Is our pursuit data shared with other ChatGPT users?
Defense Explorer only returns data for the signed-in user’s company membership. How OpenAI handles conversation content is set by your ChatGPT plan and settings; review those with your security team.
04Can ChatGPT submit a bid or email a buyer through Defense Explorer?
No. The server has no tool that submits, sends or contacts anyone. The only writes are watchlist saves, recording a decision a person made, and acknowledging an amendment.
Bring defence tender data into ChatGPT.
Start a Team trial, add the connector, and ask with sources attached.