The Agent Skills Directory

[COMMAND_EXECUTION]: The module modules/feature-extraction.md utilizes shell tools like awk, grep, and wc to analyze file metrics.
[DYNAMIC_EXECUTION]: modules/feature-extraction.md includes a Python script for sentence splitting and statistical analysis.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted markdown content into the agent's context and incorporates it into generation prompts, which can be exploited for indirect prompt injection.
Ingestion points: Local markdown files via **/*.md globs.
Boundary markers: The skill uses blockquotes as delimiters for exemplar text.
Capability inventory: Uses Bash, Read, Write, and Edit tools.
Sanitization: There is no evidence of sanitization or validation of the ingested exemplar content.

style-learner