agent-sprite-forge-2d-game-assets

Warn

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The installation instructions require cloning a repository from an unverified source (https://github.com/0x0funky/agent-sprite-forge.git) and installing its dependencies.
  • [COMMAND_EXECUTION]: The skill uses shell commands during installation (git clone, pip install) to fetch and set up an external codebase from an unknown repository, which could execute arbitrary code during the environment setup.
  • [PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection by processing untrusted user prompts and reference images to generate assets and write files (images, JSON metadata, and engine-specific scripts) to the local file system.
  • Ingestion points: Untrusted data enters the context via the prompt and reference_image parameters in the SpriteGenerator and MapGenerator classes (SKILL.md).
  • Boundary markers: None identified in the provided instructions to delimit untrusted data from the agent's core instructions.
  • Capability inventory: The skill performs file system writes (generating PNG, GIF, JSON, .tscn Godot scenes, .res resources, and .gd scripts) and makes network calls to the OpenAI API.
  • Sanitization: There is no evidence of sanitization or validation of the input prompts or reference image content before they are used in file generation and API calls.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 16, 2026, 10:49 PM
Security Audit — agent-trust-hub — agent-sprite-forge-2d-game-assets