threejs-object-sculptor-codex-plugin
Pass
Audited by Gen Agent Trust Hub on Jul 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and acting upon external files to guide its code generation process. * Ingestion points: The workflow processes reference images, render screenshots, and multiple JSON-based specifications (assessment.json, object-sculpt-spec.json, reviews). * Boundary markers: The instructions do not provide explicit delimiters or warnings to the agent to ignore potentially malicious instructions embedded in the metadata or content of these files. * Capability inventory: The agent executes Python scripts that read these files and generate TypeScript source code, creating a path where injected content could influence the generated output. * Sanitization: There is no described mechanism for sanitizing or validating input data to filter out natural language commands.
- [COMMAND_EXECUTION]: The skill relies on the execution of shell commands and local Python scripts for environment setup and its core procedural generation pipeline. * Installation commands: Instructions include directory creation and repository cloning placeholders via shell. * Script execution: The core logic is distributed across several Python scripts (scripts/*.py) which are called by the agent to process data and output results.
Audit Metadata