← Vetro Attest · Free Art. 12/14 checker · Free EU AI Act tools
Audit log for Flowise chatflows
Flowise builds chatflows visually, but it is built on top of LangChain.js: every node in a flow compiles down to a LangChain component, and running a chatflow runs a LangChain chain or agent underneath. That matters for auditing, because it means the same recording point used for hand-written LangChain code applies here too, once you can reach it.
What to record
- Every LLM call the flow makes: model, prompt, response, tokens.
- Every tool call, including Custom Tool nodes that run your own code.
- Which chatflow and which node produced each action, so a multi-node flow stays attributable end to end.
Capturing it at the LangChain layer
If you run Flowise programmatically — embedding it as a library, or running its server yourself and calling flows through code rather than only the hosted UI — you can attach the same duck-typed callback handler used in the LangChain guide. Vetro's handler needs no dependency on langchain itself:
import { Vetro } from '@vetro/core';
import { createVetroCallbackHandler } from './langchain-callback.mjs';
const run = vetro.startRun({ name: 'support-chatflow' });
const handler = createVetroCallbackHandler(run);
// pass `handler` wherever your integration exposes LangChain callbacks
Every event is hash-chained and the run is signed with ed25519.
If you only have the hosted API
Calling a chatflow through Flowise's prediction API without custom code gives you the top-level request and response, not the internal LLM and tool calls a multi-step flow makes along the way. That is real evidence, just at a coarser grain — record it as a single llm.request / llm.response pair with @vetro/core's Run API rather than assuming visibility you do not have.
Custom Tool nodes
A Flowise Custom Tool runs code you wrote. Wrap that function directly with run.toolCall() before it executes and run.toolResult() after, and check an inline policy before the call — the one place in a chatflow you have full, unambiguous control over.
Check your own logs
Paste a Vetro ledger or an OpenTelemetry GenAI trace export into the free checker: it verifies integrity and shows which record-keeping obligations your evidence covers, in your browser. When you need something to hand to an auditor or a customer, Vetro Attest issues a 29 USD counter-signed report mapping the evidence to EU AI Act articles, including a list of what the log cannot show.
See a sample ledger verified in one click, check your own logs free or get the countersigned EU AI Act report.