creating-sticker-packs
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the agent to execute a bundled Python script (scripts/process_sticker_sheet.py) using a workspace-provided Python interpreter. The script manages file system operations including directory creation, image processing, and report generation within a designated project folder.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-controlled data such as sticker names and pack titles via a JSON request file. This data is subsequently used in the generation of prompts for image creation tools and to define output filenames, creating an entry point for indirect instructions.
- Ingestion points: Project metadata including pack names and individual sticker labels are ingested from the sticker_request.json file.
- Boundary markers: The skill defines specific prompt recipes and visual contracts to limit the influence of external data on the agent's behavior.
- Capability inventory: The system allows for file read/write access, directory removal within the workspace, and complex image manipulation using standard libraries.
- Sanitization: The processing script applies normalization to names (slugification) and validates that file paths remain relative to the authorized directory to prevent path traversal.
Audit Metadata