The Agent Skills Directory

[SAFE]: The skill operates entirely within a local environment using Python's standard library. No external dependencies or packages are required or installed.
[DATA_EXPOSURE]: The skill explicitly states that no data is fetched or scraped from LinkedIn. Users are instructed to provide their own profile information manually or via JSON files. There are no network calls or remote data transmission mechanisms implemented in the scripts.
[COMMAND_EXECUTION]: The skill uses local Python scripts (headline_scorer.py, profile_completeness_auditor.py, about_section_builder.py) to process text. These scripts use argparse and re for analysis and do not utilize dangerous functions like eval(), exec(), or subprocess.run().
[INDIRECT_PROMPT_INJECTION]: While the skill processes user-provided profile data (an injection surface), the associated scripts lack the capabilities (network access, file system writes, or shell execution) required to weaponize such an injection. The risk is negligible.
[DYNAMIC_CONTEXT_INJECTION]: No dynamic execution placeholders (e.g., shell command execution at load time) were found in the documentation or instructions.

linkedin-profile