This page is machine-translated from French. Read the French original.

Agent flow

A flow is a sequence of steps defined in advance, visually constructed, and always executed in the same order. Where an agent decides what to do, a flow does what was planned.

When to prefer a feed to an agent

Take a stream when the procedure is stable and the result must be constant: a daily extraction, a standardized processing, an API call followed by formatting.

Take an agent when the approach depends on what one finds along the way.

The manufacturer

Réglages → Agents → Constructeur de flux.

The interface is a chain of blocks. Each block receives what the previous one has produced and transmits its own result to the next, via... variables.

The available blocks

Startup. Declares the input variables of the stream — what is provided to it at startup.

API call. Sends an HTTP request: URL, method, headers, request body. The response is stored in a variable. An option allows the response to be sent directly to the user, bypassing the model—useful when the result is already usable.

Instructions for the model. Give the model an instruction, applied to the available variables: summarize, extract, reformat, classify. The result is stored in a variable.

Web extraction. Retrieves the content of a URL and stores it in a variable.

The blocks are chained together as many times as necessary.

An example

Objective : Each morning, review new publications from a regulatory source and produce an internal memo.

1. Startup — variable source_url.
2. Web extraction — retrieves the content of source_url → contenu_brut.
3. Instruction — «Extract only the publications dated yesterday or today. If none, answer exactly: NONE.» → nouveautes.
4. Instruction — «Write an internal memo of no more than ten lines based on nouveautes, citing the references. » → note.
5. API call — publishes note in the internal distribution tool.

This flow, associated with a planned task, runs every morning without intervention.

Use a stream

A published feed becomes a skill It can be used by agents. Therefore, it can be called:

• from a conversation, by entrusting the task to the agent; ;
• from a scheduled task; ;
• within another sequence.

Best practices

Test block by block. A stream is built by adding one step at a time and checking its output, rather than writing the whole thing before the first attempt.

Name your variables explicitly. liste_factures_impayees better than resultat2 — the readability of the feed depends on it, and you will come back to it in six months.

Plan for the empty case. What should happen if the API returns nothing, if the page is inaccessible, if there is no new content? An explicit instruction on this case avoids absurd outcomes.

Constrain the format of the model's outputs. «"Answer only with a bulleted list, without introduction" produces a result usable by the next step; a vague instruction produces a text to be reworked.

Beware of secrets. An API call may require a key. Treat these streams with the same care as a production script, and restrict who can modify them.