discord-http-api-preview
Fail
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: HIGHPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (HIGH): This skill defines a high-privilege interaction surface with the Discord API, making it susceptible to indirect prompt injection attacks.
- Ingestion points: The skill facilitates the agent taking input from user prompts to populate API parameters (e.g.,
contentincreate_message,reasoninban_user_from_guild) and processing potentially untrusted data returned from Discord API responses. - Boundary markers: There are no boundary markers or instructions in the documentation (e.g., in
SKILL.mdor operation files) to help the agent distinguish between data and instructions when executing these API calls. - Capability inventory: Across the 227 operation files, the skill grants the agent capabilities for administrative actions including
ban_user_from_guild(references/operations/ban-user-from-guild.md),delete_channel(references/operations/delete-channel.md), andbulk_delete_messages(references/operations/bulk-delete-messages.md). - Sanitization: No input validation, escaping, or filtering mechanisms are described in the documentation, relying entirely on the agent's internal logic to handle potentially adversarial input.
Recommendations
- AI detected serious security threats
Audit Metadata