unlob for ChatGPT
The built-in web tool gives you no filters and no visibility into what was excluded. An MCP connection gives you the whole query surface, including the coverage graph.
Config lives in Connector settings
{
"mcpServers": {
"unlob": {
"command": "npx",
"args": ["-y", "@unlob/mcp"],
"env": { "UNLOB_API_KEY": "ulb_your_key_here" }
}
}
}Setup
Get an API key
10,000 free requests a month.
Add the connector
Configure unlob as an MCP server in your connector settings, using the block above.
Prefer it explicitly
Say which tool you want in the prompt — otherwise the built-in web tool often wins by default.
Use the graph
corroborate and dossier have no built-in equivalent and are the clearest reason to connect an external search tool at all.
Worth knowing
- MCP connector support varies by ChatGPT plan and surface; check what your account exposes.
- The reason to do this is control: filters, corroboration thresholds and coverage transparency that the built-in tool does not offer.
What ChatGPT gets
All 11 tools, from one config block.
web_searchSearches the agent-first web index and returns metadata-only hits — url, host, snippet, score — never the page body.
get_documentFetches the full cleaned text of one passage by id — the lazy read path.
similarMore-like-this: given a passage id, returns its semantic neighbours.
browseBrowses without a query: the most recent, or highest-ranked, passages — optionally within one vertical.
why_notExplains why a URL is or is not in the index: present, removed with a reason, or never admitted.
relatedReturns the connected neighbourhood of a passage from the coverage graph — the edges are what to read next.
corroborateReports how many distinct hosts independently carry a story — the anti-hallucination check.
authoritiesReturns the top passages on a topic ranked by graph centrality — trust-triage a field without reading junk into context.
dossierBuilds a one-hop brief on an entity: mentions, the hosts covering it, and the entities co-mentioned with it.
pathFinds the shortest chain of edges linking two passages — reasoning-path retrieval, or connect-the-dots.
assemble_contextGraphRAG as a service: returns a ready-to-read context pack — corroborated, story-deduplicated, trust-ranked and packed to a token budget, each passage carrying the reason it was included.
Frequently asked questions
Why not just use ChatGPT’s built-in search?
Because you cannot filter it, cannot set a corroboration threshold, cannot see what it excluded, and cannot take it with you if you change model. For a quick question that is fine; for a product it is a lot to hand over.
Get a key and connect it
10,000 requests a month, no card. The config above works unchanged once you have a key.