req-intake
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill automates tasks by generating and executing shell commands for
lark-cliandmeegle. It directly uses user-supplied strings for command arguments like titles and field values. - [INDIRECT_PROMPT_INJECTION]: There is an attack surface where untrusted data from a user (e.g., requirement titles) is ingested and used in sensitive operations. * Ingestion points: User input for titles, problem scenarios, and examples in SKILL.md Step 1. * Boundary markers: None present in the command templates to separate instruction from data. * Capability inventory: File creation/modification via
lark-cliand project item creation viameegle. * Sanitization: None mentioned; the skill assumes the agent safely handles the interpolation of text into shell commands.
Audit Metadata