discord
Pass
Audited by Gen Agent Trust Hub on Aug 24, 2026
Risk Level: SAFE
Full Analysis
- [CREDENTIALS_SAFE]: The skill uses environment variables (
DISCORD_TOKEN,DISCORD_USER_TOKEN) injected by the platform for authentication. It explicitly instructs the agent not to echo or print these secrets. User-provided tokens (BYOC) are managed through standard environment injection. - [COMMAND_EXECUTION]: The skill executes a local Python script (
scripts/discord_user.py) to handle Discord interactions. This script validates IDs, uses structured JSON output, and implements a safety gate (--confirmflag) that prevents the AI from sending messages without explicit user approval. - [EXTERNAL_DOWNLOADS]: The skill references the
discord.py-selflibrary but specifically forbids runtime installation viapip, requiring it to be pre-installed in the trusted sandbox environment. It uses the official Discord API domain (discord.com). - [INDIRECT_PROMPT_INJECTION]: While the skill reads external data (Discord messages), it processes them through structured JSON parsers. It does not interpolate this data into commands or executable contexts. Risks are mitigated by the fact that write actions require an additional, separate confirmation step by the user.
Audit Metadata