lark-workflow-meeting-summary

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill relies on executing system shell commands and the lark-cli binary to perform its primary functions. It uses date for time calculation and various lark-cli subcommands for authentication, searching meeting records, querying metadata, and creating documents.
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection because it processes untrusted meeting notes and metadata retrieved from an external platform.
  • Ingestion points: Data enters the agent context through lark-cli vc +notes and lark-cli drive metas batch_query commands (SKILL.md).
  • Boundary markers: There are no explicit delimiters or specific instructions for the agent to ignore or isolate instructions that might be embedded within the meeting content being summarized.
  • Capability inventory: The skill has the capability to execute shell commands via lark-cli and perform write operations to cloud documents via lark-cli docs +create and +update (SKILL.md).
  • Sanitization: No evidence of validation, escaping, or filtering of the retrieved meeting content is present before it is used in report generation or document updates.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 09:52 AM