bggg-creator-image2psd
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes a JSON manifest file to automate the assembly of images into PSD layers, which creates a vulnerability surface for indirect prompt injection if the agent is directed to process untrusted manifests.
- Ingestion points: The manifest file is parsed and processed by the
scripts/image2psd.pyscript. - Boundary markers: Absent. The script lacks validation to ensure manifest-defined file paths are restricted to the project sandbox.
- Capability inventory: The
image2psd.pyscript possesses file-read (via Pillow) and file-write (PSD, PNG, and ZIP) capabilities. - Sanitization: While standard JSON parsing is employed, there is no enforcement of path sandboxing or strict file-type whitelisting.
- [EXTERNAL_DOWNLOADS]: The documentation includes instructions for cloning the skill from the author's GitHub repository (
github.com/binggandata/bggg-skills.git) and installing standard Python packages. These resources are consistent with the vendor's own infrastructure and expected setup procedures. - [COMMAND_EXECUTION]: The skill's primary workflow involves the execution of local Python scripts (
scripts/image2psd.py,scripts/init_project.py) to manage projects and perform image-to-PSD conversions, which is aligned with the skill's documented purpose.
Audit Metadata