deck-tuner

Pass

Audited by Gen Agent Trust Hub on Apr 13, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches MTG bulk data from the Scryfall API and retrieves web content via WebFetch for metagame research. These operations are essential to the skill's stated purpose of providing up-to-date deck optimization.
  • [COMMAND_EXECUTION]: The skill relies on a suite of Python CLI tools (e.g., mana-audit, scryfall-lookup, build-deck) mapped in pyproject.toml. These tools are used for quantitative analysis and file management within the skill's working directory.
  • [DATA_EXPOSURE]: The skill processes user-provided deck lists and collection exports (CSV files). This access is scoped to the local environment and intended for hydrating the deck data with card metadata.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface in Step 4, where it fetches untrusted metagame articles from the web. While the skill follows a rigid multi-step logic and uses subagent 'Self-Grills' to verify conclusions, it lacks explicit boundary markers or sanitization instructions for the ingested web content.
  • [SAFE]: The skill does not contain hardcoded credentials, obfuscated code, persistence mechanisms, or unauthorized privilege escalation. The configuration and instructions are professional and well-documented.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 13, 2026, 03:53 AM