meitu-visual-me

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Recommends installing the meitu-cli tool from the official npm registry to enable the visual assistant capabilities.- [COMMAND_EXECUTION]: Utilizes the local meitu CLI to execute visual generation workflows and performs standard file operations like mkdir and mv for output management.- [DATA_EXFILTRATION]: Accesses vendor credentials in ~/.meitu/credentials.json and transmits excerpts of local context files such as USER.md and MEMORY.md to the Meitu OpenAPI as part of the personalized prompt generation process.- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple sources:
  • Ingestion points: Reads local files including USER.md, MEMORY.md, and PROFILE.md, and ingests user-provided images.
  • Boundary markers: The prompt construction logic weaves external data naturally into narrative sentences without explicit delimiters.
  • Capability inventory: The agent can invoke the meitu CLI for network operations and perform local file writes.
  • Sanitization: No specific sanitization or filtering of the extracted local content is performed before interpolation into prompts.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 07:19 AM