The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: Fetches and executes an installation script from a Google-owned domain (antigravity.google) and pipes it to the shell. This script downloads the Antigravity CLI binary and verifies its integrity using SHA-512.
[COMMAND_EXECUTION]: Uses os.execv within a Python broker script to launch the agy binary in a pseudo-terminal (PTY), facilitating headless OAuth authentication in environments without a TUI.
[COMMAND_EXECUTION]: Orchestrates the agy CLI by executing shell commands with user-provided tasks to delegate work to a secondary AI agent harness.
[PROMPT_INJECTION]: Exhibits an indirect prompt injection surface as the skill is designed to ingest and incorporate output from an external agent harness (Google's Antigravity) into the primary agent's context.
Ingestion points: Captures output (stdout) from the agy CLI command execution.
Boundary markers: The skill does not define specific delimiters or instructions to treat the sub-agent output as untrusted.
Capability inventory: The execution environment allows shell command execution, network operations via the CLI, and file system access for token management.
Sanitization: No sanitization or validation of the sub-agent's output is performed before it is integrated into the agent's workflow.

invoking-antigravity