Automation Workflow Audit Tool — n8n & Make

Paste any n8n workflow export or Make (Integromat) blueprint and receive a structured, AI-powered audit report in seconds.

What you get

• Executive summary — what the workflow does and its overall risk level
• Step-by-step flow analysis — each node explained in plain language
• Risk analysis — missing error handlers, fragile triggers, external dependency risks
• Concrete issues — specific weaknesses and ambiguities flagged directly
• Improvement recommendations — architecture, stability, monitoring
• Quality score — 1–10 with explanation
• Operations & maintenance advice
• Optional test plan — happy path, failure, duplicate trigger, credential issues

Output is Markdown, pushed to the Actor dataset and ready to copy into Notion, Confluence, or any docs tool.

---

Input

Field	Required	Description
platform	✅	n8n or make
workflowJson	✅	Exported JSON from n8n or Make
model	—	claude-sonnet-4-5 (default, recommended) or claude-haiku-4-5 (faster, cheaper)
includeTestPlan	—	true (default) — append a test checklist to the report
language	—	de, en, fr, es — auto-detected from workflow content if not set

---

Quickstart

1. Export your workflow

n8n: Open your workflow → ⋯ menu → Download → saves as .json

Make: Open your scenario → ⋯ menu → Export Blueprint → saves as .json

2. Run the Actor

Paste the JSON into the Workflow JSON field in the Apify UI, select your platform, and hit Start.

3. Retrieve the report

The audit report is saved in two places after each run:

After each run, three files are available under Storage → Key-Value Store:

File	Format	Best for
report.pdf	PDF	Sending to clients, attaching to tickets
report.docx	Word	Editing, adding company branding
report.md	Markdown	Pasting into Notion, Confluence, GitHub

Go to the run page → Storage → Key-Value Store → click the file → Download.

The full audit is also available as a JSON record in Storage → Dataset (documentation field) for programmatic access.

---

Example input (UI / JSON mode)

{
  "platform": "n8n",
  "workflowJson": "{\"name\":\"Invoice → Slack Notification\",\"nodes\":[{\"name\":\"Webhook\",\"type\":\"n8n-nodes-base.webhook\"},{\"name\":\"Parse Invoice\",\"type\":\"n8n-nodes-base.function\"},{\"name\":\"Send to Slack\",\"type\":\"n8n-nodes-base.slack\"}],\"connections\":{}}",
  "model": "claude-sonnet-4-5",
  "includeTestPlan": true,
  "language": "en"
}

Example output (excerpt)

# Workflow Audit: Invoice → Slack Notification

## Executive Summary
This workflow receives invoice data via a webhook, processes it in a
function node, and sends a Slack notification. Overall risk level: **medium**.
No error handler is present — a failed Slack call will silently drop the event.

## Quality Score
**4 / 10** — The core logic is sound but the workflow lacks error handling,
retry logic, and any form of logging. Not production-ready as-is.
...

---

Environment variables

Set in Actor → Settings → Environment variables:

Variable	Required	Description
ANTHROPIC_API_KEY	✅	Your Anthropic API key — get one at console.anthropic.com

---

Limits

• Maximum workflow JSON size: 500 KB
• Supported platforms: n8n, Make (Integromat)
• Supported output languages: DE, EN, FR, ES (auto-detected or manual)

---

Local development

# Install dependencies
npm install

# Type-check only
npm run type-check

# Run with tsx (no build step)
ANTHROPIC_API_KEY=sk-... npm run dev

# Build TypeScript → dist/
npm run build

# Run compiled output
npm start

Deploy to Apify

npm install -g apify-cli
apify login
apify push --force