The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches official documentation and resource indices from the neon.tech domain using curl to ensure the agent has access to current information.
[REMOTE_CODE_EXECUTION]: Instructions are provided for installing the official Neon CLI via a remote script (https://neon.tech/install.sh | bash) and for setting up local environments using npx neon init. These are standard installation methods for the service.
[COMMAND_EXECUTION]: The skill guides the agent to perform management tasks using the Neon CLI (neonctl), curl for API interactions, and various ORM-specific CLI tools (e.g., drizzle-kit).
[INDIRECT_PROMPT_INJECTION]: The skill provides a mechanism to ingest external documentation fetched at runtime, which represents a potential attack surface.
Ingestion points: Documentation URLs and the llms.txt index fetched via curl in SKILL.md and references/referencing-docs.md.
Boundary markers: None explicitly defined in the fetch commands.
Capability inventory: The agent is authorized to execute shell commands (curl, neonctl), write to local configuration files (.env), and perform network operations.
Sanitization: No specific sanitization or filtering of the fetched documentation is mentioned.
[SAFE]: Recommendations for secret management, such as using .env files and environment variables instead of hardcoding credentials, align with industry security standards.

using-neon