discord-bot-members

Warn

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • Data Exposure & Exfiltration (HIGH): The skill is designed for mass data harvesting of Discord server members, specifically mentioning scales of 100,000+ members. This data (usernames, nicknames, roles) is stored in local YAML files, creating a significant repository of PII (Personally Identifiable Information) that could be targeted for exfiltration.
  • Indirect Prompt Injection (MEDIUM): 1. Ingestion points: Discord API member profiles (usernames, nicknames). 2. Boundary markers: None mentioned in instructions or output schema. 3. Capability inventory: Local file system writes and network access to Discord API. 4. Sanitization: None mentioned. If downstream tools or agents process these member lists, malicious strings in usernames or nicknames could trigger unintended behavior.
  • Unverifiable Dependencies (MEDIUM): The documentation explicitly mentions 'discord.py-self'. This library is a known fork for self-bots (automating user accounts), which is a high-risk activity often associated with scraping and account bans. Its inclusion in a 'bot token' skill is suspicious and suggests non-standard automation practices.
  • Command Execution (LOW): The skill requires the agent to execute local Python scripts with various flags, which is the primary mechanism for the data harvesting operation.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 16, 2026, 01:31 PM