Other connections FAQ
Common questions about what ctx| connects today and what is on the roadmap.
Can I connect Confluence / Notion / Google Drive?
Not as a built-in, first-class connector in the open-source documentation today. The supported path is Git-first: put canonical prose in the repo (or sync there) so ingestion and agents see the same truth as engineers.
Does ctx| scan my laptop?
No. ctx| indexes repositories you register (and data derived from them in your deployment). It does not install a generic endpoint agent on developer machines as part of the core stack described here.
How do private repositories work?
You provide credentials acceptable to your Git host (HTTPS tokens, SSH deploy keys, or GitHub App installation flows). The codesearch service uses those when cloning; exact UX depends on hosted vs self-hosted configuration.
Can one user belong to multiple organisations?
Yes. Users can be members of many orgs; the active organisation is carried on
the session, and URLs use /:orgSlug/. MCP requires an explicit orgSlug query
parameter. See Managing your organisation.
Where do API keys and OAuth fit in?
Programmatic access uses Bearer API keys or OAuth for MCP and related flows. See Connecting tools and Authentication.
What if ingestion is slow or fails?
Large monorepos and cold indexes take longer. Failures are often auth, network,
or resource limits on the codesearch / Zoekt layer. Check service logs and
repository indexReady (or equivalent) status in the API.