The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute local Python scripts using shell commands to extract text from documents (extract-text.py), normalize markdown formatting (normalize-markdown.py), and launch a local HTTP server (serve.py) to provide a UI for the resume editor.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from user-provided resumes and job descriptions. This represents an ingestion surface where a malicious document could contain hidden instructions aimed at biasing the agent's analysis or bypassing constraints, though the risk is mitigated by the specific nature of the task and the use of dedicated parsing libraries.
Ingestion points: Document file paths passed to extract-text.py and raw text pasted into the agent context.
Boundary markers: The agent is instructed to use markdown blocks and structured logic, but there are no explicit boundary markers or sanitization of the content before it is processed by the AI.
Capability inventory: The skill can execute local Python scripts via shell commands and initiate a local network server.
Sanitization: Document parsing is handled by the pdfminer.six library and standard XML parsers, which provide a layer of abstraction from raw file vulnerabilities.
[EXTERNAL_DOWNLOADS]: The skill specifies a dependency on the pdfminer.six Python package and includes several bundled well-known JavaScript libraries (Sortable, html2canvas, and marked) to support the functionality of the local web editor.

resume-matcher