The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to download and execute the @nomadamas/k-skill package from the NPM registry. This package is a vendor-owned utility provided by the skill's author (nomadamas) to manage instruction sets and script execution.
[COMMAND_EXECUTION]: Shell commands are executed via npx to trigger internal Python scripts and retrieve configuration data. These operations are scoped to the skill's primary functionality of job matching.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of user-provided resumes to extract keywords for search queries.
Ingestion points: Resume text is ingested via the resume_text parameter or by reading a local file through the resume_file argument.
Boundary markers: The instructions include "Hard boundaries" and "Hard rules" that explicitly prohibit the agent from processing sensitive identifiers (SSNs, exact addresses) or performing automated actions like job applications without user consent.
Capability inventory: The skill utilizes network access (urllib.request) and local file system access (reading resumes). It does not have write access to sensitive system directories.
Sanitization: Input data is sanitized through keyword extraction and regex-based filtering within the job_posting_match.py script before being incorporated into search URLs.

job-posting-match