lark-task
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes task summaries, descriptions, and comments, which could potentially contain malicious instructions from external parties aimed at the agent.
- Ingestion points: Data enters the context via retrieval tools such as
+search,+get-my-tasks, and+get-related-tasksdefined in the reference files. - Boundary markers: The instructions do not prescribe the use of delimiters or explicit 'ignore embedded instructions' warnings when the agent reads or displays task content.
- Capability inventory: The skill has significant write capabilities including task creation, modification, and file attachment uploading via
lark-cli. - Sanitization: There is no mention of sanitization or character escaping for content retrieved from external task fields before it is processed by the agent.
Audit Metadata