add-mesh-surface-sampled-particles

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources which could be manipulated to influence agent behavior.
  • Ingestion points: Target repository instructions and component data retrieved from the threenix MCP tool (SKILL.md).
  • Boundary markers: No specific delimiters or safety instructions are defined to separate injected content from existing scene code.
  • Capability inventory: The skill possesses file system write access, the ability to install npm/yarn packages, and the capability to execute shell commands for building/testing the project.
  • Sanitization: The instructions direct the agent to write the retrieved content "verbatim," providing no mechanism for sanitizing or validating the code before integration.
  • [EXTERNAL_DOWNLOADS]: The skill is designed to fetch code and assets from a remote MCP provider and write them directly to the user's filesystem.
  • [DYNAMIC_EXECUTION]: The workflow involves writing local script files based on remote tool output and subsequently executing build or type-check processes on those files.
  • [COMMAND_EXECUTION]: The skill executes shell commands to install dependencies and perform build operations, which are triggered by the results of external tool calls.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 08:11 AM
Security Audit — agent-trust-hub — add-mesh-surface-sampled-particles