punk-ip-article-illustrations
Pass
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill ingests untrusted external data, such as article text, Markdown files, or web pages, and incorporates it into prompts for an image generation tool. It lacks explicit boundary markers or isolation instructions to prevent the agent from obeying instructions embedded within that external data.
- Ingestion points: User-provided article content, local files, and external web links as described in
SKILL.mdandreferences/article-workflow.md. - Boundary markers: None; the skill does not specify delimiters or warnings to ignore embedded instructions in the processed articles.
- Capability inventory: The skill uses a local script for file management, writes character assets and illustrations to the local disk, and can modify existing Markdown files if requested.
- Sanitization: Minimal; only character slugs are validated using a regular expression in the
scripts/character_registry.pyscript. - [COMMAND_EXECUTION]: The skill involves the execution of a local Python script,
scripts/character_registry.py, provided with the skill. This script is used to manage character assets and manifest files in the user's home directory (~/.punk-ip-assets). The script performs standard file operations such as copying images and writing JSON metadata.
Audit Metadata