gws-workflow-meeting-prep

Pass

Audited by Gen Agent Trust Hub on Mar 20, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted data from external sources. \n
  • Ingestion points: Event descriptions, attendee lists, and agenda details from Google Calendar retrieved via the gws tool in SKILL.md. \n
  • Boundary markers: Absent; the skill does not use specific delimiters or instructions to isolate retrieved meeting content from agent instructions. \n
  • Capability inventory: The skill uses the gws command-line interface to read data; it does not demonstrate capabilities for file modification or unsolicited network transmission. \n
  • Sanitization: Absent; there is no evidence of content filtering or validation for the retrieved calendar data. \n- [COMMAND_EXECUTION]: The skill executes the gws binary to fetch calendar events, which is the primary intended function of the workflow. \n- [NO_CODE]: No executable scripts are distributed with the skill; it operates by defining metadata and calling an existing CLI tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 20, 2026, 04:56 PM