Skills
Navigate
skills/samuraigpt/generative-media-skills/muapi-social-pack/Gen Agent Trust Hub

muapi-social-pack

Pass

Audited by Gen Agent Trust Hub on Sep 30, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts user-defined text in the caption_idea input and interpolates it into a prompt for the nano-banana-2-edit image model.
  • Ingestion points: User input provided via the caption_idea parameter in SKILL.md.
  • Boundary markers: The instructions use single quotes to delimit the user input (e.g., 'caption: '<caption_idea>'.'), which is a weak boundary for preventing prompt injection.
  • Capability inventory: The skill utilizes muapi CLI commands and curl for network-based image processing.
  • Sanitization: The skill does not implement any validation or escaping of the user-provided text before inserting it into the prompt.
  • [COMMAND_EXECUTION]: The execution steps require the agent to run local shell commands using the muapi CLI tool.
  • Commands used include muapi image edit, muapi auth configure, and muapi predict wait.
  • The agent is also instructed to use curl for direct API interaction if CLI aliases are unavailable.
  • [EXTERNAL_DOWNLOADS]: The skill performs network communication with the domain api.muapi.ai to send image data and receive processed results.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 30, 2026, 07:17 AM