The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx zeabur@latest, which downloads the latest version of the vendor's command-line interface from the npm registry during execution.
[COMMAND_EXECUTION]: The skill executes various shell commands using the Zeabur CLI to perform infrastructure management tasks, such as creating, listing, and deleting environment variables.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes data from .env files and user-provided key-value pairs.
Ingestion points: Reads content from local .env files via the -f flag and accepts variable values from user input via the -k flag in SKILL.md.
Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are present for the variable values being processed.
Capability inventory: The skill uses subprocess calls to execute the zeabur CLI which interacts with the vendor's cloud API.
Sanitization: The skill identifies risks related to shell expansion (${VAR}) and CLI parser limitations (Cobra's StringToStringVar), recommending the use of single quotes or the web dashboard for sensitive or complex values.

zeabur-variables