Guides

Put your agents to work. Practical walkthroughs for building teams, connecting tools, and keeping your projects organized.

Orchestrating agents

Build mixed-harness teams with roles, model and effort selection, visible workers, and repeatable review loops.

A hands-on AgentGrid tutorial with a Claude Code builder, a separate Codex reviewer, and a small pagination bug you can reproduce.

How AgentGrid reports Ollama Cloud usage and protects your Chrome session.

Run Claude Code, Codex, OpenCode, Antigravity, Cursor, Devin, Grok, Kimi, and Pi on one canvas.

Real PTY-backed terminals with native input, safe paste, search, and canvas-aware gestures.

Chromium browser panes with per-workspace session isolation and a programmable surface for QA workers.

Markdown note panes with Whisper dictation and inline Mermaid diagrams.

Talk to an agent pane over a real-time voice call, configure voice preferences, and review call history.

Create an AgentGrid account, sign in to the desktop app, reset a password, and manage a Free or Pro plan.

Filter panes on the canvas by type, live status, and session activity — non-matching panes dim, layout stays put.

Pair two AgentGrid desktops and open the same project on both — panes, terminals, notes, and layout stay in live two-way sync.

The Expo/React Native companion that pairs to your desktop so you can watch and prompt agents from your phone.

The master's MCP tool surface, the browser_qa Playwright integration, and project-local role overrides.