steam-publish

Pass

Audited by Gen Agent Trust Hub on Oct 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides operational guidance for Steam publishing using official tooling (steamcmd) provided by a well-known service (Valve/Steam). References to external documentation point to the official partner.steamgames.com domain.
  • [DATA_EXPOSURE]: While the skill involves the use of sensitive credentials (Steam account details and login tokens), it includes robust security warnings. It specifically advises against committing passwords or tokens to source control and provides a detailed workflow for using GitHub Secrets and cached login tokens (config.vdf) to maintain security in automated environments.
  • [COMMAND_EXECUTION]: The shell commands provided (e.g., steamcmd execution) are standard for the tool's intended use case. The scripts utilize placeholders for sensitive information, prompting the user to provide their own values rather than executing hardcoded or obfuscated commands.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes VDF (Valve KeyValues) files, which are a form of structured data. The instructions focus on creating these files from templates rather than parsing untrusted external data, significantly reducing the risk of injection through data ingestion surfaces.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 8, 2026, 09:07 PM