The Agent Skills Directory

[SAFE]: The skill provides legitimate architectural guidance, code samples, and configuration patterns for the Neon Functions platform. The provided examples follow security best practices, such as recommending JWT authentication for public endpoints and cross-isolate fan-out using Postgres LISTEN/NOTIFY.
[COMMAND_EXECUTION]: Documentation includes standard administrative and development commands for the neon CLI (e.g., neon dev, neon deploy, neon env pull). These tools are used for project lifecycle management and environment synchronization.
[EXTERNAL_DOWNLOADS]: The skill utilizes well-known and trusted Node.js packages from established technology providers, including Neon Database (@neon/config, @neon/env), Vercel (ai), Hono (hono), and Anthropic (@modelcontextprotocol/sdk). All external references target official documentation or repositories.
[CREDENTIALS_UNSAFE]: The skill demonstrates secure secret management by advising users to inject credentials through environment variables or secure deployment flags (--env) rather than hardcoding them. It also provides detailed patterns for verifying short-lived JWTs to protect sensitive operations.
[INDIRECT_PROMPT_INJECTION]: As the skill involves building AI agents, it explicitly identifies the risk of processing untrusted data and provides remediation guidance, such as input validation, sanitization, and least-privilege tool access.

neon-functions