browser-read-x
PassAudited by ClawScan on May 1, 2026.
Overview
This skill appears to do what it says—run a bundled browser-page extractor to read the currently open X/Twitter content—with no evidence of exfiltration, persistence, or destructive behavior.
This looks reasonable for its stated purpose. Before installing, be aware that it runs JavaScript in the active browser page and returns visible page content, including content from logged-in or protected X/Twitter pages if you use it there.
Findings (2)
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 can run this extractor on the currently open page and read the page's visible content.
The skill intentionally runs bundled JavaScript in the active browser page. This is disclosed and purpose-aligned for extracting page content, but users should understand that code executes in the page context.
Run `browser act` with `kind="evaluate"` and pass the contents of `skills/browser-read-x/extract.js` as `fn`.
Invoke it only on pages whose visible content you are comfortable sharing with the agent, and keep the skill source trusted and reviewed.
Private or protected X/Twitter content that is open in the browser could be returned to the agent as extracted text.
The skill is designed to work on pages that may be accessible because the user is already authenticated in the browser. It does not request credentials, but it can extract content visible through the user's current browser access.
`web_fetch` is noisy or blocked (auth-required/public X/X-protected pages).
Do not run it on private, protected, or sensitive pages unless you want that content included in the agent conversation.