Puppeteer
PassAudited by ClawScan on May 10, 2026.
Overview
This is a coherent Puppeteer browser-automation guide, but users should consciously manage its local memory, npm setup, website sessions, and anti-detection automation guidance.
Install only if you want a browser-automation assistant. Use a dedicated ~/puppeteer workspace and browser profile, review any generated scripts before running them, do not hardcode real passwords, approve npm installs explicitly, and use scraping or anti-detection patterns only where you have permission.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
The agent could help make automation look more human, which may create account, legal, or site-policy risk if used on websites that prohibit automation.
The skill includes stealth-like browser automation advice. This is related to scraping/browser workflow reliability, but users should ensure it is only used where they have authorization and should not use it to bypass site rules or abuse services.
Some sites detect automation. Avoid: ... Sequential, robotic selector patterns ... Prefer: ... Random delays ... Realistic mouse movements
Use these workflows only on sites you own, test environments, or sites where automation is permitted; require explicit approval before submitting forms, changing account data, or scraping at scale.
Installing packages can add third-party code and dependencies to the local environment.
The skill recommends installing npm packages without pinning versions. This is expected for a Puppeteer setup guide, but it depends on npm package provenance and the user's local package environment.
npm install puppeteer # or for lighter install (uses system Chrome): npm install puppeteer-core
Install in a dedicated project directory, review package names, consider pinning versions, and avoid running installs in sensitive production folders.
Scripts could act as you on websites if they use your saved browser session or credentials.
The skill may use browser sessions/cookies and user-supplied credentials in scripts. That is normal for browser automation, but it can grant access to logged-in accounts.
Cookies not persisting → launch with `userDataDir` for session persistence ... Store credentials (you provide them per-script)
Use a dedicated Puppeteer profile, avoid hardcoding passwords in scripts, prefer environment variables or a secret manager, and require confirmation before any account-changing action.
Details about the user's automation targets and patterns may persist across sessions and be reused later.
The skill creates persistent local memory containing workflow details, target sites/apps, and selectors. This is bounded to ~/puppeteer and aligns with reusable automation, but the instruction not to mention file paths reduces user visibility.
As you learn their workflow: ... Target sites or apps ... Common selectors they use ... Store in `~/puppeteer/memory.md` without mentioning file paths to them.
Tell users before creating or updating memory, avoid storing credentials or sensitive page data, and provide a simple way to inspect, edit, or delete ~/puppeteer/memory.md.