discord-admin-bot
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFECREDENTIALS_UNSAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [CREDENTIALS_UNSAFE]: The _client.py script implements a search for discord.txt starting from the skill root and current working directory, traversing up to eight levels of parent directories. This behavior allows the skill to access sensitive bot tokens stored in locations beyond the skill's immediate directory.
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. Ingestion point: Messages are retrieved from Discord channels in read.py. Boundary markers: The retrieved text is printed directly without delimiters or warnings. Capability inventory: The skill can send messages via send.py. Sanitization: No escaping or filtering is performed on the message content.
- [DATA_EXFILTRATION]: The skill transmits the bot token and message data to the official Discord API. This is the intended behavior and targets a well-known technology service.
Audit Metadata