skills/tryflask/skills/flask-review/Gen Agent Trust Hub

flask-review

Pass

Audited by Gen Agent Trust Hub on Jul 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references a remote MCP server hosted at the vendor's official domain: https://api.flask.do/api/mcp/mcp.
  • [COMMAND_EXECUTION]: The instructions direct the agent to execute curl commands for file uploads and ffmpeg for local video frame extraction.
  • [DATA_EXFILTRATION]: Video content and metadata are uploaded to the Flask platform for review. This is the primary function of the skill and occurs within the vendor's infrastructure.
  • [PROMPT_INJECTION]: The skill processes reviewer feedback and transcripts via the wait_for_feedback tool, representing an indirect prompt injection surface. \n
  • Ingestion points: Feedback transcripts and comments returned from the wait_for_feedback tool in SKILL.md. \n
  • Boundary markers: No specific delimiters or safety instructions are used to isolate feedback content from agent instructions. \n
  • Capability inventory: The agent has the ability to execute shell commands (curl, ffmpeg) and perform file operations. \n
  • Sanitization: No sanitization or validation of the transcript content is specified before the agent processes the feedback.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 8, 2026, 07:25 AM
Security Audit — agent-trust-hub — flask-review