The Agent Skills Directory

[COMMAND_EXECUTION]: The script feishu_doc.py uses subprocess.run to invoke the feishu-docx CLI for exporting and importing document content.
Evidence: Found in cmd_read, cmd_export_wiki, and cmd_import_wechat within feishu_doc.py.
Observation: The implementation uses argument lists rather than shell strings, which is a secure practice to prevent shell injection.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from external Feishu documents and chat rooms, which could contain instructions intended to influence the agent's behavior.
Ingestion points: feishu_doc.py fetches external data via cmd_read, cmd_list_blocks, and cmd_read_chat.
Boundary markers: Absent. Data is passed directly to the agent's context without clear delimiters or warnings to ignore embedded instructions.
Capability inventory: feishu_doc.py contains capabilities for shell command execution (via feishu-docx), writing/overwriting documents, deleting content, and modifying sharing permissions (e.g., cmd_permission for making documents public).
Sanitization: Absent. The skill does not perform validation or escaping of the content fetched from Feishu APIs before presenting it to the agent.

feishu-doc