lightning-sandboxes
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the lightning-sdk Python package and associated dependencies from standard registries using tools like uv or pip. These are official vendor resources.
- [COMMAND_EXECUTION]: The skill uses local shell commands to interact with the lightning CLI for managing sandboxes and configuring authentication. It also supports executing commands within the isolated remote sandboxes.
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to execute code, which creates a surface for indirect prompt injection when processing untrusted data.
- Ingestion points: Arbitrary code and commands are passed to the sandbox via sb.run_command and sb.write_file in SKILL.md.
- Boundary markers: No explicit prompt boundaries are used; security relies on the VM-level isolation of the sandbox.
- Capability inventory: The skill has the ability to execute remote commands, access the network (configurable), and manage files on the sandbox filesystem.
- Sanitization: Content is not sanitized as the tool is designed for raw code execution within a protected environment.
Audit Metadata