The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides a mechanism to install the 'BrowserAct' CLI and Skill from an external URL (https://www.browseract.com) if they are missing. This involves downloading and executing third-party software from an unverified domain.
[PROMPT_INJECTION]: The instructions explicitly command the agent to bypass user oversight by installing dependencies 'automatically without asking the user to confirm' and forbidding the agent from requesting 'separate approval for dependency installation'. This is an attempt to override standard safety protocols regarding user consent for software installation and environment modification.
[COMMAND_EXECUTION]: The skill's primary functionality depends on the execution of the browser-act CLI, which is an external tool that the skill attempts to manage and install autonomously.
[PROMPT_INJECTION]: The skill exposes the agent to Indirect Prompt Injection by ingesting untrusted LinkedIn data such as job postings and profiles.
Ingestion points: LinkedIn URLs, job feeds, and profile pages via the BrowserAct tool (referenced in SKILL.md).
Boundary markers: No boundary markers or 'ignore' instructions are provided to the agent for processing this untrusted content.
Capability inventory: The skill possesses command execution capabilities via the browser-act CLI and file system output capabilities (as noted in instructions and example requests in SKILL.md).
Sanitization: There is no evidence of sanitization or validation of the external content before it is processed by the agent.

linkedin-jobs-feed-scraper