discord
Pass
Audited by Gen Agent Trust Hub on Jul 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill restricts its operations to the 'message' tool with a specific Discord channel, providing a controlled environment for agent activities.
- [SAFE]: Authentication is handled through a required configuration field for the Discord token, ensuring that secrets are managed securely by the platform rather than being hardcoded or exposed.
- [SAFE]: The skill includes examples for accessing files for media uploads (e.g., 'file:///tmp/example.png'), which is a standard functional capability for a messaging tool and does not target sensitive system paths.
- [PROMPT_INJECTION]: The skill enables reading content from Discord channels, creating a surface for indirect prompt injection where the agent may process untrusted data. Ingestion points: 'read' action defined in SKILL.md. Boundary markers: None defined in the skill instructions. Capability inventory: Subprocess calls for Discord actions such as send, edit, delete, and presence updates. Sanitization: No explicit content sanitization or instruction to ignore embedded commands is provided.
Audit Metadata