The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automates web interactions using form_input, computer, and read_page tools. It includes specialized logic for different Applicant Tracking Systems (ATS), such as extracting iframe tokens using javascript_tool for Greenhouse. These operations are restricted to the browser context and gated by user approval.
[PROMPT_INJECTION]: The skill is subject to indirect prompt injection as it ingests untrusted data from external job postings (Step 1 and 3). This risk is mitigated by a mandatory human-in-the-loop approval step (Step 6) where the user reviews all proposed answers before any data is filled, and a final confirmation (Step 8) before submission. Ingestion points include job URLs in SKILL.md, while the primary capability inventory consists of browser automation and local logging.
[EXTERNAL_DOWNLOADS]: The skill connects to external job boards. These are well-known technology services (Lever, Greenhouse, Workday) and are necessary for the skill's core functionality.

apply