The Agent Skills Directory

[COMMAND_EXECUTION] (LOW): The skill utilizes the Bash tool to execute local Python helper scripts (fetch_docpath.py and validate_apis.py). While these are internal project tools, they are invoked with arguments (such as <docPath> URLs) derived from user or external input, which poses a minor risk of command injection if input is not handled securely by the scripts.
[EXTERNAL_DOWNLOADS] (LOW): The fetch_docpath.py script is designed to download content from the Feishu/Lark Open Platform documentation URLs. While the destination is likely restricted to a specific domain, downloading external content for processing is a known attack vector.
[PROMPT_INJECTION] (LOW): Indirect Prompt Injection Surface Detected.
Ingestion points: External API documentation fetched via URL or parsed from HTML files (references/README.md, SKILL.md).
Boundary markers: Absent. The instructions do not define delimiters or warnings to ignore instructions embedded within the fetched documentation content.
Capability inventory: The agent has permissions to edit files (Edit), run local commands (Bash), and read the filesystem (Read, Grep, Glob).
Sanitization: Absent. There is no requirement mentioned to sanitize or validate the content retrieved from external URLs before the LLM processes it for code generation.

openlark-api