The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external data in the form of Job Descriptions (JDs) and Resumes. This content is used by the agent to generate predicted interview questions and model answers.
Ingestion points: The prompts/jd-driven-prep.md file defines steps to upload a JD and a resume for analysis.
Boundary markers: The skill relies on natural language instructions to follow specific frameworks like STAR, CAR, and SOAR, which acts as a structural constraint.
Capability inventory: The skill is capable of reading files from sibling directories (../job-description-skill/) and writing HTML reports to the user's local Desktop.
Sanitization: The instructions do not explicitly detail sanitization methods for the ingested text, relying on the LLM's internal safety filters.
[COMMAND_EXECUTION]: The skill instructs the agent to write files to a specific local path (~/Desktop/Claude skills/) to provide the user with a prepared interview report. While this is a documented feature intended for the user, it involves interacting with the local file system based on instructions derived from external data.

bq-skill