Skip to main content
You don’t need to build a dashboard to get an answer from Enfors. Connect an AI client to the Enfors MCP server and ask questions in plain English — Enfors queries the decision ledger for you and can render a shareable report.

​
How it works

When you ask a question, the connected client calls Enfors MCP tools to fetch exactly the data it needs, reasons over the results, and can present them as a branded view via render_ui. Enfors never asks the model to write SQL — the tools expose safe, parameterized queries.
Natural-language query in an AI client

Asking Enfors from Claude Code / Cursor

​
Connect a client

1

Add the Enfors MCP server

Point your client at the Enfors MCP endpoint and authorize with OAuth / SSO. See MCP overview for the endpoint and per-client setup.
2

Ask a question

For example: “show me what the team produced in the last three days.”
3

Explore and follow up

Drill in, test hypotheses, and refine — all in conversation.
4

Render and share

Ask for a report and Enfors renders a branded view (PII removed) you can share.

​
Example questions

  • “Show me what the team produced in the last three days.”
  • “Which agents cost the most this week, and what did they deliver?”
  • “Find sessions flagged as rework and summarize why.”
  • “Compare two people’s success rate and spend this month.”
  • “What was unusual about session sess_…?”

​
Good to know

It queries the ledger directly

No SQL, no separate BI tool — answers come straight from your live data.

PII is removed

Generated reports have personal data scrubbed. See Privacy.
Next: the full MCP tool reference.