The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted text from external Markdown files, which represents an indirect prompt injection surface.
Ingestion points: External Markdown content provided via file paths or direct input as described in SKILL.md.
Boundary markers: Absent; the skill does not use specific delimiters or instructions to prevent the model from obeying instructions embedded within the text to be translated.
Capability inventory: The skill is capable of reading and writing to the local file system and executing bundled Python scripts (SKILL.md).
Sanitization: None; the content is passed to the model for translation as-is.
[COMMAND_EXECUTION]: The skill executes local Python scripts, scripts/chunk_markdown.py and scripts/resolve_config.py, to handle file chunking and configuration loading. These scripts use Python standard libraries and perform mechanical text processing without executing arbitrary shell commands or using dangerous functions like eval() or exec() on untrusted input.
[EXTERNAL_DOWNLOADS]: The skill lists PyYAML as an optional dependency for parsing configuration files. This is a well-known package from a standard registry. No other external or unverified code downloads are performed at runtime.

soia-pkm-translate-article-zh