The Agent Skills Directory

[NO_CODE]: The skill references eight JavaScript files in a src/ directory (such as detectUnfollowers.js and monitorAccount.js) that are not included in the skill bundle, making it impossible to verify the security of the actual logic being executed.
[COMMAND_EXECUTION]: The instructions direct users to paste scripts directly into the browser developer console while active on X.com. This is a high-risk activity that bypasses standard browser security boundaries and allows scripts full access to the user's authenticated session, including cookies and private account information.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from X.com profiles. 1. Ingestion points: Scrapes follower lists and profiles from X.com. 2. Boundary markers: None provided to distinguish between account data and instructions. 3. Capability inventory: The scripts use localStorage for data persistence and have the capability to download files to the user's system. 4. Sanitization: There is no evidence of sanitization or validation of the external content before it is processed or stored.

follower-monitoring