The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill provides a bash snippet that uses python3 -c to encode user-supplied emoji strings. This involves direct string interpolation into a shell command, which presents a potential command injection vulnerability if the input contains malicious characters intended to escape the Python string literal.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted user-supplied content and transmit it to external Discord channels, creating a surface for indirect prompt injection.
Ingestion points: User-supplied message text, embed descriptions, and emoji labels.
Boundary markers: The skill instructions emphasize showing a full preview and obtaining explicit user confirmation before any API call is executed.
Capability inventory: The skill utilizes the Bash tool to execute curl for network requests to Discord and to perform simple string processing.
Sanitization: There is no explicit logic provided in the instructions for sanitizing or escaping the content within the generated JSON payloads or shell commands.
[SAFE]: API credentials (webhooks and bot tokens) are loaded from standard local environment files (.env, .env.local), which is a secure practice for secret management.
[SAFE]: All network activity is restricted to discord.com, which is a well-known service appropriate for the skill's stated purpose.

discord-bot