The Agent Skills Directory

[SAFE]: The provided Python scripts (brand_voice_analyzer.py and seo_optimizer.py) implement path traversal protection using Path.cwd().resolve() and relative_to checks. They also enforce a 1 MiB file size limit on processed inputs to prevent resource exhaustion.
[COMMAND_EXECUTION]: The skill uses local Python scripts to perform lexical and SEO analysis. These commands are executed locally and are restricted to the skill's working directory.
[INDIRECT_PROMPT_INJECTION]: The skill has an ingestion surface as it processes external text files for content analysis.
Ingestion points: scripts/brand_voice_analyzer.py and scripts/seo_optimizer.py take a user-supplied file path as an argument.
Boundary markers: The instructions explicitly state to "Review any private text before sharing the output" and limit execution to the working directory.
Capability inventory: The scripts are limited to reading text, performing regex-based word counts, and printing diagnostics to standard output. They do not interpret data as code or perform network requests.
Sanitization: Path resolution logic prevents access to files outside the permitted directory, and the processing is purely lexical.

content-creator