proxy-printer

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted user-supplied data in the form of deck lists and Moxfield links, establishing a surface for indirect prompt injection attacks.\n
  • Ingestion points: The parse-deck command accepts user-supplied files or links to process them into the internal deck schema.\n
  • Boundary markers: SKILL.md defines an 'Iron Rule' that prohibits the agent from using its own training data or user-supplied oracle text, mandating that all rendered content come from a verified Scryfall lookup.\n
  • Capability inventory: The skill utilizes CLI tools authorized to write to the local file system (PDF generation and art caching) and perform network requests to external domains.\n
  • Sanitization: Input strings used for file lookups are converted into alphanumeric 'slugs' to prevent path manipulation attacks.\n- [EXTERNAL_DOWNLOADS]: The fetch-art and download-mtgjson tools download card metadata and ASCII art from external domains, including asciiart.eu and asciiart.website.\n
  • Description: These downloads are contextually appropriate for the skill's primary function of providing card proxies and attributed artwork but target non-whitelisted third-party sources.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 04:35 PM
Security Audit — agent-trust-hub — proxy-printer