The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted data from external sources, including job description text, URLs, and job portal content. This creates a surface for indirect prompt injection where malicious instructions embedded in a job description could attempt to influence the agent's output during CV generation or email drafting.
Ingestion points: The skill processes user-supplied job description text, JD URLs, and data from external job portals and application forms (e.g., in the 'apply' and 'scan' modes).
Boundary markers: There are no explicit instructions or delimiters provided to the agent to help it distinguish between its core instructions and the potentially untrusted content it is processing.
Capability inventory: The skill instructs the agent to perform sensitive tasks such as generating CVs (PDF, LaTeX, Text), drafting outreach emails, and providing answers for job applications.
Sanitization: The instructions do not include any guidance on sanitizing, escaping, or validating the input data retrieved from external job descriptions or web portals.
[NO_CODE]: The skill is a 'docs-only' implementation. It contains markdown-based instructions and a routing logic table but does not bundle any executable scripts, binaries, or automated installation processes. It relies entirely on the agent's native capabilities or the user's manual setup of the upstream repository.

career-ops