dingtalk-calendar

Pass

Audited by Gen Agent Trust Hub on Aug 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes Python scripts (calendar_free_slot_finder.py, calendar_schedule_meeting.py, and calendar_today_agenda.py) to interface with the dws command-line utility. These scripts employ subprocess.run with argument lists to execute commands securely, which is a recommended practice to prevent shell injection vulnerabilities.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided inputs such as meeting titles, descriptions, and participant names, which creates a standard attack surface for indirect prompt injection. The risk is managed by using dedicated scripts that perform data normalization and use specific tools to resolve identities.
  • Ingestion points: User-supplied meeting details (titles, descriptions, names) processed by the scripts and SKILL.md instructions.
  • Boundary markers: The skill does not explicitly use boundary markers for user-supplied strings within its prompts.
  • Capability inventory: The skill can execute the dws binary to read, create, and delete calendar events and meeting room bookings across all Python scripts.
  • Sanitization: calendar_schedule_meeting.py performs time normalization using the datetime library, and the skill resolves participant names to userIds via a dedicated search command (dws aisearch) before use.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 29, 2026, 05:05 PM
Security Audit — agent-trust-hub — dingtalk-calendar