discord-bot-architect
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The provided templates for both JavaScript and Python implement standard patterns for dynamically loading bot commands and extensions from the local filesystem.
- In
sub-skills/discordjs-v14-foundation.md, the script usesrequire(filePath)inside a loop to load command and event handlers from specific subdirectories. - In
sub-skills/pycord-bot-foundation.md, the script usesbot.load_extension()to dynamically load Python modules from the./cogsdirectory. - [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of bots that process untrusted data from the Discord API, creating an inherent attack surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent/bot context via Discord interactions, including slash command options, button clicks, and select menu choices as seen in
sub-skills/interactive-components-pattern.mdandsub-skills/pycord-bot-foundation.md. - Boundary markers: The boilerplate templates do not implement specific delimiters or 'ignore' instructions for user-supplied content.
- Capability inventory: The resulting bots have network access (Discord Gateway/API) and filesystem access (for loading local scripts).
- Sanitization: The examples demonstrate functional handling but do not include explicit security sanitization or validation of user-provided strings before they are processed or echoed back.
Audit Metadata