Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent newsletter automation skill, but a broadcast-authentication flaw could let a misconfigured public endpoint send email to every confirmed subscriber.
Review before installing. Do not activate the broadcast workflow until NEWSLETTER_SECRET is set to a strong random value and the public fallback is removed; put the webhooks behind authentication or rate limiting, sanitize subscriber-supplied fields before email rendering, add signup abuse controls, and fix the drip state so welcome emails are sent only once. Also publish a privacy/consent notice and restrict access to the subscriber Sheet.
workflows/03-broadcast-sender.json:23Broadcast Authentication Bypass Through a Public Fallback Secret
workflows/01-subscriber-signup.json:23Confirmation Tokens Generated With a Non-Cryptographic PRNG
workflows/01-subscriber-signup.json:5Public Signup Webhook Can Be Used for Unsolicited Email and Resource Abuse
Thanks for subscribing to our newsletter.
Please confirm your subscription by clicking the link below:
If you didn't subscribe, you can safely ignore this email.
", "options": {} } } ``` ### Technical Analysis A newsletter signup endpoint must normally be public, but the workflow provides no compensating abuse controls. There is no CAPTCHA, signed-form verification, IP or address-based rate limit, replay prevention, cooldown, request-size limit, or robust email schema validation. The only validation is whether the supplied value contains `@`. Every accepted request can write a record to Google Sheets and trigger an SMTP email. Because storage uses `appendOrUpdate` with the email address as the matching field, repeated requests can also replace pending subscriber data and confirmation tokens. The network operation itself is necessary for double opt-in, but allowing unlimited anonymous callers to invoke it excee ...[truncated 1010 chars]workflows/01-subscriber-signup.json:23Subscriber-Controlled Values Are Embedded Into HTML Email Without Escaping
Thanks for subscribing to our newsletter.
``` The welcome workflow uses the stored name directly in HTML templates: ```js welcome: { subject: `Welcome aboard, ${item.name}!`, body: `Hi ${item.name},
Thanks for confirming your subscription. Here's what to expect:
Stay tuned — your first tips are coming in a few days.
` }, tips: { subject: `3 tips to get you started, ${item.name}`, body: `Hi ${item.name},
Here are 3 tips to get the most from your subscription:
More great content coming your way soon!
` } ``` Stored names and broadcast content are embedded directly into broadcast HTML: ```htmlHi {{ $json.name }},
{{ $json.content ...[truncated 2812 chars]workflows/02-welcome-sequence.json:44Day-Zero Drip State Causes Repeated Welcome Emails
The skill describes collecting subscriber emails, names, sources, confirmation status, tokens, and analytics in Google Sheets, but it does not prominently warn users that personal data will be stored and used for automated campaigns and reporting. This creates a privacy and consent risk because operators may deploy the workflow without adequately informing subscribers or configuring appropriate data handling safeguards.
Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.
## What It Does
1. **Double Opt-In Signup** — Webhook receives signups, validates email, sends confirmation link, stores in Sheets
2. **Welcome Drip Sequence** — Automatically sends Day 0 (welcome), Day 3 (tips), Day 7 (resources) emails
3. **Broadcast Sender** — API-triggered broadcast to all confirmed subscribers with unsubscribe links
4. **Daily Analytics** — Subscriber counts, growth metrics, confirmation rates, top sources
The broadcast feature is presented as a normal capability, but the documentation does not clearly warn that invoking it will send mass email to all confirmed subscribers. Without an explicit caution, an operator could trigger large-scale messaging unintentionally or without appropriate review, creating spam, reputational, and compliance risks.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
curl -X POST https://your-n8n.com/webhook/newsletter/signup \
-H "Content-Type: application/json" \
-d '{"email": "test@example.com", "name": "Test User", "source": "api-test"}'
This is a JSON manifest/workflow file, so vague-trigger checks apply. The path "newsletter/signup" indicates activation on a generic signup POST endpoint, but the file provides no explicit constraints, negative examples, or scoping details about who may invoke it or under what contexts, which can make activation expectations ambiguous.
This workflow automatically sends email to subscriber addresses on a schedule without any built-in user-facing disclosure, approval checkpoint, or operational safeguard in the workflow itself. In a newsletter skill this behavior is expected, but if the subscriber sheet is populated incorrectly or abused, the system can send unsolicited messages at scale, creating spam, compliance, and reputation risk.
This is a manifest/config file, so vague-trigger checks apply. The activation path "newsletter/broadcast" identifies a high-level action but provides no explicit constraints, negative examples, or narrower invocation context, which can make the skill easier to invoke unintentionally in systems that route by descriptive trigger names.
This workflow sends emails to subscriber addresses and includes personalized content, which is a safety-relevant network operation involving user data. In this file there is no confirmation step, user-visible logging, or inline warning/comment disclosing that subscriber email addresses and message content will be transmitted via SMTP.
The workflow hard-codes the subject and HTML message in English. Under the natural-language policy check, forcing a specific language without offering user choice or documenting a justified locale constraint is a policy violation.
This is a manifest-style JSON file, so vague-trigger checks apply. The schedule is defined only as a 24-hour interval, but the file does not specify when it should run, under what conditions it should be disabled, or any limiting context, which can make invocation scope overly broad.
SQP-3 only covers natural-language language or locale policy violations. The content here does not show a language/locale restriction, so this is only marginally applicable; however, there is no actual locale policy issue visible in the file.
No suspicious patterns detected.