SquadCue creates AI agent mission control with approval inbox
TL;DR. SquadCue introduces a local-first mission control system for AI CLI agents, turning them into "employees" with chat, memory, and an approval inbox. - The system supervises AI agents running on a local machine, allowing users to resume sessions and manage tool calls. - A core feature is the gated chat, requiring human approval for risky actions like shell commands or file writes. - SquadCue integrates with Claude Code and offers a visual canvas for orchestrating repeatable agent pipelines.
- SquadCue offers local-first "mission control" for AI CLI agents, treating them as employees.
- The system allows users to chat with, resume, and orchestrate AI agent sessions, with memory persistence.
- A key feature is the human-in-the-loop approval inbox for risky tool actions like shell commands or file writes.
- It supports Claude Code and can adapt to other CLIs, running as a lightweight FastAPI server with a single-file UI.