The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill implements dynamic loading of code components from local directories to initialize bot commands and events. While this is a standard pattern for the discord.js and py-cord frameworks, it involves executing code from computed file paths at runtime.
Evidence: In src/index.js, the script iterates over the commands and events directories using fs.readdirSync and imports each file using require(filePath).
Evidence: In main.py, the script iterates over the ./cogs directory and uses bot.load_extension(f"cogs.{filename[:-3]}") to dynamically load Python modules.
[INDIRECT_PROMPT_INJECTION]: The skill provides templates for handling user interactions, which introduces a surface where untrusted data from Discord users is ingested into the bot's execution context.
Ingestion points: The interaction object in the JavaScript ping command and the ctx / message options in the Python main.py slash command definitions.
Boundary markers: The templates do not implement explicit delimiters or instructional boundaries to isolate user-supplied strings from the bot's logic.
Capability inventory: The provided examples are limited to responding within the Discord channel (interaction.reply, ctx.respond) and do not demonstrate high-risk capabilities such as file system writing or subprocess execution.
Sanitization: No input sanitization or validation logic is present in the boilerplate code for user-supplied strings.

discord-bot-architect