The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of CLI tools for environment setup and infrastructure management.
Executes npx neon@latest init --agent for project bootstrapping and tool installation.
Uses neon link, neon checkout, and neon deploy for branch-first development workflows and infrastructure-as-code management.
Utilizes the skills CLI to dynamically add other functional skills from the vendor's repository.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download of configuration and documentation from trusted vendor sources.
Downloads agent skills from the official neondatabase/agent-skills GitHub repository.
Fetches live documentation in markdown format from https://neon.com/docs/ to ensure current technical accuracy.
Installs the Neon MCP server from https://mcp.neon.tech/mcp using npx add-mcp.
[INDIRECT_PROMPT_INJECTION]: The skill architecture includes fetching external documentation (llms.txt and .md files) to supplement its knowledge base. While ingesting external data is a known vector for indirect prompt injection, the skill targets official vendor-controlled domains (neon.com), which significantly mitigates this risk in a production environment.
[CREDENTIALS_UNSAFE]: The skill instructions discuss the management of sensitive environment variables such as NEON_API_KEY and DATABASE_URL. It promotes standard security best practices by directing these values into git-ignored .env or .env.local files rather than hardcoding them or passing them as cleartext arguments.

neon