hugging-face

Pass

Audited by Gen Agent Trust Hub on Apr 28, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing the @membranehq/cli package via NPM. This tool is provided by the vendor to facilitate interaction with their platform.
  • [COMMAND_EXECUTION]: Several shell commands using the membrane CLI are used for authentication and task execution. This is the intended method for using the skill.
  • [PROMPT_INJECTION]: The skill ingests untrusted data from Hugging Face (e.g., repository files and discussions) through actions like list-repository-files and get-discussion in SKILL.md. While this represents an indirect prompt injection surface, the skill relies on the Membrane platform for processing and does not perform unsafe local operations with the data. No explicit boundary markers or local sanitization are defined.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 28, 2026, 10:02 AM