lark-minutes
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
lark-clicommand-line interface to interact with external APIs. This is the intended functionality of the skill for retrieving meeting information. - [INDIRECT_PROMPT_INJECTION]: The skill processes Lark/Feishu meeting URLs to extract tokens used in API calls. While this presents a theoretical attack surface where malicious data could be processed, it is restricted to meeting identifiers.
- Ingestion points: Feishu/Lark URLs provided in the skill instructions.
- Boundary markers: None specific to input sanitization are defined in the file.
- Capability inventory: Execution of
lark-cliforminutesandvcresources. - Sanitization: The instructions focus on extracting specific token formats, which acts as a rudimentary filter.
Audit Metadata