AI services ยท CONNECTIONBuild and maintain hosted MCP context servers in MCP Studio: servers, sources, tools, retrieval rules, analytics and usage.
Explore 23 workflow actions for AI Context by MCP Studio, including account permissions and operation limits.
Choose an action in the workflow builder, connect an account, and map its inputs. These actions are available in the catalog; your account access and a successful test determine whether they work for your setup.
Add a public source (website, docs, GitHub, API, MCP server or PDF) to a server and queue it for indexing.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Replace a server's tools, or add and remove some; 3 to 10 must remain.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Add a plain-language rule for how a server, or one source, retrieves content.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Create a hosted MCP server from sources, queue them for indexing, and get its endpoint and client config.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Delete a custom tool and detach it from every server.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Delete a retrieval rule.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Soft-delete a server; it can be restored from the dashboard for 7 days.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Read a server's request volume, success rate, latency, and tool, source and client breakdowns.
Read a custom tool and its parameters.
Read the trial state, analytics tier and source limits.
Read a server's indexing queue; keep polling while status is crawling.
Read a server with each source's indexing progress, its endpoint and client config.
Read the plan, MCP request usage, and server and source limits.
List the account's custom tools.
List a server's MCP requests, by time range, tool, client and outcome (paid analytics).
List a server's retrieval rules.
List a server's tools, the built-in catalog and the custom tools that can be attached.
List the servers the key can reach, including ones deleted in the last 7 days.
Switch a server between public and private. A token issued on first going private is not returned here; create one on the server's page.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Queue a re-index of every source, or one, so the server picks up changed content.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Stop sending a credential to a remote MCP source, keeping the source.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Remove a source by ID or URL and delete its indexed content; a source not on the server gives removed 0.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Pause, resume or reorder a retrieval rule.
A repeated request may repeat the external action. Review the previous result before retrying a write.
Use the account and permissions required for your process. Review the connection check before configuring a workflow action.
Select a listed operation and map its required inputs. Use test records and add a human approval before writes that need review.
Inspect step outputs and confirm the result in AI Context by MCP Studio. A saved workflow is not a completed provider action.
