The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from job descriptions and resumes to provide scoring and rewrite suggestions. A malicious job description could attempt to influence agent behavior during the tailoring process.
Ingestion points: Data enters via user-provided text files jd.txt and resume.txt which are processed by the skill's script.
Boundary markers: The instructions lack explicit delimitation or "ignore instructions" wrappers for the external content when the agent performs subsequent rewrites.
Capability inventory: The skill uses a local script for file reading and data analysis. The agent uses the script's output to perform text generation.
Sanitization: The Python script tokenizes and filters text for keyword matching, but does not sanitize the content for adversarial prompt patterns that might be interpreted by the LLM during the rewrite workflow.
[COMMAND_EXECUTION]: The skill utilizes a local script scripts/resume_matcher.py for its core analysis functionality.
The script uses only Python standard library modules (argparse, json, re, pathlib) and does not perform network operations or unexpected file modifications.
The script reads local files provided as command-line arguments, which is standard for the described resume analysis functionality.

resume-tailor