feishu-doc-reader
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- [NO_CODE]: The skill documentation references several essential functional scripts including scripts/read_doc.sh, scripts/get_blocks.sh, scripts/read_feishu_doc.py, scripts/get_feishu_doc_blocks.py, and scripts/test_auth.py. However, none of these files were included in the provided file set for analysis.
- [PROMPT_INJECTION]: The skill is designed to fetch and process data from external documents and spreadsheets, which represents a surface for indirect prompt injection attacks where malicious content in a document could attempt to hijack the agent's behavior.
- Ingestion points: Content is retrieved from the Feishu Open API (open.feishu.cn) and brought into the agent's active context (SKILL.md).
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions to separate untrusted document content from system prompts.
- Capability inventory: The skill uses python3 and curl to fetch data and process it, with the ability to read local configurations and document content (SKILL.md).
- Sanitization: There is no mention of sanitization or validation of the ingested document content to prevent the execution of embedded instructions.
Audit Metadata