The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a local Python script scripts/yomiyasu_lint.py to perform static analysis and quality checks on the text. The script uses standard libraries and does not perform network operations or access sensitive system files.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user-provided text or file content as its primary input for rewriting. While this presents an ingestion surface for potentially malicious instructions hidden in data, the skill's core instructions emphasize strict semantic preservation and it lacks capabilities (such as external network access) that would allow for successful data exfiltration or escalation from such an injection.
Ingestion points: User-provided text or file paths specified in the skill arguments.
Boundary markers: The skill instructs the agent to analyze the 'claims' and 'meaning' of the source text before rewriting, which serves as a logical check against simple instruction overrides.
Capability inventory: Shell execution of the internal linting script.
Sanitization: No explicit sanitization of input text is performed before processing.

yomiyasu