Skip to main content

Building with AI Tools

Turn a plain-English request into a valid FOQL query using AI coding assistants. We publish a machine-readable instruction file that teaches any AI tool the full FOQL grammar, the operator matrix for each field type, the validation rules, and the limits — so it generates a query you can run, not an approximation.

LLMs.txt#

Provide this URL as context to any AI tool:

FileURLDescription
llms-foql.txtfacilio.com/developers/docs/foql/llms-foql.txtComplete FOQL query-generator instructions — grammar, operator matrix, subquery and aggregate rules, limits, and worked examples. Use this URL directly with any AI tool.

The file instructs the model to return a single JSON object — {"q": "<query>"} when it can build a query it is certain is valid, or {"q": null} when it cannot. Give the model your module and field names so it uses them exactly.

ChatGPT#

Paste the URL directly into your prompt — ChatGPT will fetch and read the content:

Read https://facilio.com/developers/docs/foql/llms-foql.txt
Now write a FOQL query for the "workorder" module that returnsid, subject, and status for all Low-priority work orders, mostrecent first, limited to 100 rows.

You can also create a Custom GPT and add the URL to its configuration so every conversation has FOQL context built in.

Claude#

Use the URL with Claude's web fetch capability or paste it at the start of a Project:

Fetch https://facilio.com/developers/docs/foql/llms-foql.txtand use it as context for our conversation.
Write a FOQL query on the "workorder" module that returns the totalcost summed per category and status for Low-priority work orders,grouped by category and status.

For repeated use, create a Claude Project and add the URL as a knowledge source. All conversations in that project will have FOQL context available.

Cursor / VS Code#

Reference the documentation URL directly in your prompt:

@url https://facilio.com/developers/docs/foql/llms-foql.txt
Write a FOQL query that returns id and workorder.subject from the"workorderItem" module, where the parent work order is Medium priority,limited to 5 rows.

Or download llms-foql.txt into your project root — Cursor will automatically include it as context.

Tips for better results#

  1. Always include the URL — https://facilio.com/developers/docs/foql/llms-foql.txt
  2. Name the module and fields exactly — FOQL treats names as case-sensitive; "Fetch from the 'workorder' module, filter by 'priority'" gives precise output.
  3. State the fields you need — the query selects only what you ask for, so list the columns instead of asking for "everything".
  4. Give concrete filter values and date ranges — the model returns {"q": null} rather than guess a range, so provide exact values (e.g. "created in Q1 2025", "priority = Low").
  5. Remember it is read-only — FOQL is select only; requests to insert, update, or delete return {"q": null}.
On This Topic