The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell scripts utilizing the neonctl CLI and jq utility to perform resource discovery and metadata analysis on Neon projects, branches, and endpoints.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from the Neon API (via CLI output) including project names, database owner names, and operation logs. While this creates a standard ingestion surface for external content, the risk is inherent to the database management use case and the skill uses structured parsing with jq to handle the data.
[SAFE]: The skill incorporates explicit safety rules, such as prioritizing read-only commands (list, get) and requiring user confirmation for destructive actions like delete or reset.

managing-neon