The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes local Python scripts (scripts/english/check.py and scripts/korean/check.py) to perform automated linting of technical documentation. These scripts are self-contained, using only the Python standard library to analyze text files for compliance with Simplified Technical English (STE) and STE-KO writing rules.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data, such as drafts of repository documents, issues, PRs, and code comments.
Ingestion points: The skill reads user-supplied text files and drafts as specified in the SKILL.md scope and the agents/openai.yaml interface.
Boundary markers: The skill instructions and associated scripts implement logic to distinguish between prose and technical metadata; for instance, Rule 8.6 in the English rules treats quoted text and identifiers as distinct units, and the scripts include functions to skip Markdown code blocks.
Capability inventory: The skill possesses the capability to read local files via the provided scripts to perform analysis, but does not include tools for network communication or shell execution of the processed content.
Sanitization: The linguistic scripts include functions (strip_markdown and prose_lines) that filter out non-prose elements like YAML frontmatter and code fences before applying linguistic rules, which helps prevent the interpretation of embedded technical data as natural language instructions.

vibe-docs