discord

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The skill/connector’s required runtime tools can ingest Discord-authored message text via the listChannelMessages and getMessage scripts, which call Discord API endpoints that return content fields from server channels/threads without any further selection gating beyond user-provided channel_id/message_id.

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). The string in references/discord-api-gotchas.md is a high-entropy, three-part token that matches Discord bot token format (Authorization: Bot ). It is a literal, non-placeholder value present in the docs and therefore appears to be a real credential (not a generic placeholder like "xxx" or "YOUR_API_KEY"). I did not flag any simple placeholders (e.g. DISCORD_BOT_TOKEN=xxx) elsewhere because those are clearly placeholders by the rules.