feishu-cli-work
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates office automation by executing
feishu-clisubcommands to interact with Feishu APIs.\n- [EXTERNAL_DOWNLOADS]: Documentation points to the author's GitHub repository (github.com/riba2534/feishu-cli) for tool setup, which is considered a vendor resource.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from the Feishu environment, posing a low-severity risk of indirect instruction injection.\n - Ingestion points: Calendar event details, task summaries, and OKR progress reports fetched from Feishu APIs via feishu-cli.\n
- Boundary markers: The skill lacks explicit delimiters or "ignore embedded instructions" warnings for the untrusted data in its instructions.\n
- Capability inventory: The skill uses
Bash(feishu-cli:*)andBash(jq:*)for platform interaction and data transformation.\n - Sanitization: Uses
jqto parse JSON API responses into structured fields, which provides basic structural validation.
Audit Metadata