enclave-manager

Configure Latch session sandboxes with proxy routing and credential injection.

2|Updated Feb 26, 2026
One-click install
npx skills add https://github.com/latchagent/latch-core --skill enclave-manager
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: enclave-manager
Source: https://github.com/latchagent/latch-core/tree/main/.agents/skills/enclave-manager
Command: npx skills add https://github.com/latchagent/latch-core --skill enclave-manager

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires docker, seatbelt, bubblewrap, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the need for secure and configurable sandbox environments in Latch sessions, providing a controlled and isolated workspace for AI agents.

Core Features & Use Cases

  • Enclave Sandboxing: Manages the lifecycle and environment configuration of sandboxes for Latch sessions.
  • Proxy Routing: Routes all traffic through the Latch Proxy to ensure secure communication.
  • Backend Detection: Detects and configures available sandbox backends (Docker, Seatbelt, Bubblewrap).
  • Security Hardening: Implements security measures like preventing shell history leaks and CA cert trust.
  • Credential Injection: Injests credentials securely into the sandbox environment.
  • Use Case: When setting up a new Latch session and needing to ensure that the environment is secure, and credentials are properly managed.

Quick Start

Use the enclave-manager skill to create a new sandbox environment for your session with the command 'enclave-manager create --sessionId <your_session_id>'.

Frequently Asked Questions about enclave-manager

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I set up a secure sandbox environment for AI agent sessions?

Setting up a secure sandbox environment for AI agent sessions involves managing isolated workspaces and routing traffic through a secure proxy. This ensures controlled communication and prevents unauthorized access during session configuration.

What is credential injection in sandboxed environments?

Credential injection in sandboxed environments securely ingests and configures authentication credentials within an isolated workspace. This prevents shell history leaks and ensures the AI agent operates with proper authorization safely.

Does Latch sandboxing work with Docker, Seatbelt, and Bubblewrap?

Latch sandboxing works with Docker, Seatbelt, and Bubblewrap by detecting and configuring these available backends. This allows flexible deployment across different isolation technologies based on your system environment.

How do I route traffic through a proxy in a Latch session?

Routing traffic through a proxy in a Latch session involves configuring the sandbox to channel all network communication through the Latch Proxy. This ensures secure data transmission and protects the isolated environment.

Why does my sandbox backend fail to detect during session setup?

Sandbox backend detection fails when Docker, Seatbelt, or Bubblewrap are not installed or configured correctly. The system requires these specific tools to be available to establish the isolated environment properly.