Using AI Leash

AI Leash is a desktop workspace for working with coding agents while keeping your project, tools, and decisions in view. Start a conversation for a project, choose a local or hosted model—or an external agent such as Claude Code or GitHub Copilot—and review the work as it happens.

The guides below explain what each part of the app does and how to use it:

• The AI Leash workspace: switch projects and conversations, manage side-panel tabs, and resize the workspace.
• Project boards: plan work as tasks on a Kanban board per project, alongside a project-wide chat.
• Agent chat: start conversations, select a provider or external agent, work with sub-agents, and use chat controls.
• Tools and permissions: understand what an agent can do and review or approve consequential actions.
• Conversation history: find, organize, resume, or delete saved conversations and sub-agent work.
• Browse and edit files: use the file tree and side-panel editor while you work with an agent.
• Terminal: run your own commands and watch ongoing work.
• Actions: save and manage project commands you use often.
• Instructions, memory, and skills: give agents durable project guidance and specialized workflows.
• Crash logging
• Debug mode

All conversation history and crash logs stay on the computer running AI Leash. In the default Ask permission mode, the app asks before an agent edits files, runs commands, or takes another consequential action.