The Agent Skills Directory

[Environment Variable Management]: The skill provides detailed guidance on configuring sensitive credentials like SLACK_BOT_TOKEN and SLACK_SIGNING_SECRET. It follows industry standards by instructing users to use .env files, which are explicitly excluded from version control via .gitignore templates. Additionally, it recommends Vercel AI Gateway to minimize the need for local AI provider keys.
[Credential Safety Patterns]: For AWS integrations, the skill correctly advises against hardcoding credentials, instead promoting Vercel's OIDC (OpenID Connect) mechanism (@vercel/functions/oidc). This is a high-security practice that reduces the risk of credential leakage.
[External Resource Integration]: The skill references and downloads dependencies from well-known and trusted ecosystems, including the Vercel AI SDK, Next.js, and the Chat SDK. These resources are standard in modern web development and used here for their intended project scaffolding purposes.
[Indirect Prompt Injection Surface]: As the skill facilitates the creation of Slack bots, the resulting application will process untrusted input from Slack messages. While the instructions provide patterns for piping this data into AI models, users should implement their own boundary markers and sanitization logic when implementing the core agent logic to mitigate potential prompt injection from end-users.

slack-agent