The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches skill definition files and updates from the vendor's official GitHub repository (github.com/PPIO/ppio-skills).
[REMOTE_CODE_EXECUTION]: Downloads and executes the NodeSource setup script (deb.nodesource.com) to configure the Node.js environment on Linux systems.
[COMMAND_EXECUTION]: Installs the ppio-sandbox-cli package globally via the NPM registry to enable sandbox management capabilities.
[PRIVILEGE_ESCALATION]: Requests administrative privileges via sudo to perform system-level package installation during the Node.js setup process.
[CREDENTIALS_UNSAFE]: Manages platform authentication using ppio-sandbox-cli auth login and requires the PPIO_API_KEY environment variable for agent invocation.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill processes user-provided payloads via the agent invoke command and reads ppio.Dockerfile during template builds in SKILL.md.
Boundary markers: No specific delimitation or 'ignore instructions' warnings are provided for the payload data passed to agents.
Capability inventory: The skill can execute arbitrary CLI commands for sandbox lifecycle management, including network operations and container deployment.
Sanitization: No explicit sanitization or validation of the input payloads or Dockerfile contents is described.

ppio-sandbox