GD Project a small lab of AI-native apps made with Godot

How-to

Learn it by doing

Short guides, one goal each. Every step has what you do, what you should see, and a screenshot of the real app, captured automatically on a fictional workspace. The agents' screens are real Claude Code and Codex screens, recorded once and replayed. For how things work underneath, see the Docs.

Your first five minutesStart a Claude Code session in a project, give it a task, and follow it from the list without watching it.5 min · 5 steps Answer without opening a terminalApprove a command or answer a question from the list, while you keep reading something else.3 min · 4 steps Focus, Grid and BoardSee one agent, four at once, or a live tile for every session, and move between them with the keyboard.3 min · 4 steps Themes and shortcutsPick a look among seven themes and change any shortcut.2 min · 2 steps Review a diff and send commentsRead what an agent changed, file by file, comment on lines, and send your comments to it in one go.4 min · 4 steps Resume, fork and find any conversationFind an old session in the History, pick up its conversation where it stopped, or try another way on a copy.4 min · 4 steps Connect a repository (its own SSH key)Give a GitHub or Azure DevOps repository an SSH key made and checked by Agent Manager, so sessions push without your own keys.5 min · 4 steps Talk to your agentsDictate with F8 into a box you send with Enter, and drive the app with F9: approve, go to a session, ask who waits.4 min · 4 steps Let an agent start helpers (MCP)Have a lead agent split a job and start helper sessions of its own, which you can watch and take over.4 min · 3 steps Follow a session from your phoneKeep an eye on a Claude Code session, and answer it, from claude.ai/code or the Claude app while you are away.2 min · 2 steps A session that documents itself in the BrainStart an agent that reads where the project stands in your Second Brain, then proposes your words, its decisions and a journal as it works.4 min · 4 steps Run an agent in a cloud sandboxGive an agent a throw-away machine on Fly.io, see what it costs as it runs, and keep a budget that only pauses.6 min · 2 steps

Start5 min5 stepsReal screens

Your first five minutes

Start a Claude Code session in a project, give it a task, and follow it from the list without watching it.

You need
  • Agent Manager GD installed, with Claude Code or Codex already installed and signed in (Install)
  1. Click + New session at the top of the list (or press Ctrl+Shift+N). Give it a name, pick Claude Code, a group and the project's folder.

    The New session dialog. Below the folder, what the session gets: its own git branch, the Second Brain, a journal in the Brain, Remote Control. The defaults are fine to start.

    Your first five minutes, step 1 of 5: The New session dialog. Below the folder, what the session gets: its own git branch, the Second Brain, a journal in the Brain, Remote Control. The defaults are fine to start.
  2. Click Start.

    The session opens in Focus view: Claude Code itself, in a real terminal, with your own login and settings. Its card in the list says ready.

    Your first five minutes, step 2 of 5: The session opens in Focus view: Claude Code itself, in a real terminal, with your own login and settings. Its card in the list says ready.
  3. Type what you want in the terminal, as you always do, and press Enter.

    Its state turns to working, in the header and on its card. You can leave it: the list keeps telling you where it is.

    Your first five minutes, step 3 of 5: Its state turns to working, in the header and on its card. You can leave it: the list keeps telling you where it is.
  4. Work on something else, in another session. When an agent needs you, you see it without looking for it.

    The card moves up into They wait for you, in orange for an OK, in yellow for a question, with what the agent asks. A toast says it too.

    Your first five minutes, step 4 of 5: The card moves up into They wait for you, in orange for an OK, in yellow for a question, with what the agent asks. A toast says it too.
  5. When it is finished, its card says done.

    Close the app if you like: the session lives on in its own terminal server. Open the app again and every agent is where you left it.

    Your first five minutes, step 5 of 5: Close the app if you like: the session lives on in its own terminal server. Open the app again and every agent is where you left it.
If it does not work
  • Claude Code is not found. Open ◐ › Setup check: it says what is missing and the command to run yourself (Setup check).
  • The state stays on ready while the agent works. States come from the agent's own hooks, which Agent Manager installs for that session only. A session started before an update needs Restart.
  • The text looks wrong in the terminal. That is a bug: tell us. The terminal must be impeccable (accents, colours, emoji, mouse).

More in the Docs: Sessions and states · The terminal

Start3 min4 stepsReal screens

Answer without opening a terminal

Approve a command or answer a question from the list, while you keep reading something else.

You need
  1. Look at the top of the list.

    They wait for you holds every session that needs you, first come first: here, one waits for your OK to run a command, another asks a question.

    Answer without opening a terminal, step 1 of 4: They wait for you holds every session that needs you, first come first: here, one waits for your OK to run a command, another asks a question.
  2. Click Approve on the one that waits for your OK (or press Ctrl+Shift+A for the first one).

    Agent Manager presses the agent's highlighted choice for you (its Yes). The card leaves the queue and the session is working again.

    Answer without opening a terminal, step 2 of 4: Agent Manager presses the agent's highlighted choice for you (its Yes). The card leaves the queue and the session is working again.
  3. On the one with a question, click Reply….

    A small box: write your answer and press Enter. It goes to that agent's terminal as if you had typed it there.

    Answer without opening a terminal, step 3 of 4: A small box: write your answer and press Enter. It goes to that agent's terminal as if you had typed it there.
  4. Want to read before you answer? Click Open.

    The session opens in Focus view on its question: here, a choice between three ways to identify a client, which you answer with the arrow keys.

    Answer without opening a terminal, step 4 of 4: The session opens in Focus view on its question: here, a choice between three ways to identify a client, which you answer with the arrow keys.
If it does not work
  • Approve picked the wrong choice. Approve presses Enter, the agent's highlighted choice. When you want another one, use Open and choose yourself.
  • A session waits but is not in the queue. Only agents with hooks report a state (Claude Code, Codex). A shell has no state.

More in the Docs: Live states · Shortcuts

Start3 min4 stepsReal screens

Focus, Grid and Board

See one agent, four at once, or a live tile for every session, and move between them with the keyboard.

  1. Focus (Ctrl+Shift+F): one session, full size.

    The terminal takes the whole space; the header says its tool, folder, branch, whether it shares the Second Brain, and its state.

    Focus, Grid and Board, step 1 of 4: The terminal takes the whole space; the header says its tool, folder, branch, whether it shares the Second Brain, and its state.
  2. Grid (Ctrl+Shift+G): up to four sessions side by side.

    Each terminal is real and live: type in any of them. Its title bar shows its state. Choose which ones from the list.

    Focus, Grid and Board, step 2 of 4: Each terminal is real and live: type in any of them. Its title bar shows its state. Choose which ones from the list.
  3. Board (Ctrl+Shift+B): a tile per session.

    A live preview of each terminal and its state. Click a tile to open it in Focus.

    Focus, Grid and Board, step 3 of 4: A live preview of each terminal and its state. Click a tile to open it in Focus.
  4. Ctrl+Shift+↓ and Ctrl+Shift+↑ go to the next and the previous session; Ctrl+Shift+K opens the command palette.

    Type a few letters of any action or session name: gr finds Grid view and the sessions in a group.

    Focus, Grid and Board, step 4 of 4: Type a few letters of any action or session name: gr finds Grid view and the sessions in a group.
If it does not work
  • A shortcut does nothing in a terminal. In a terminal, plain Ctrl keys belong to its programs: add Shift. You can change any shortcut in ◐ › Shortcuts.

More in the Docs: Views · Shortcuts

Start2 min2 stepsApp only

Themes and shortcuts

Pick a look among seven themes and change any shortcut.

  1. Click View ▾ at the top (it reads Focus ▾ or Grid ▾) and pick a theme, or type theme in the command palette.

    Seven themes, the same as Second Brain's, each shown in its own colours. The window changes at once and the panel stays open, so you can try them one after the other. Interface and terminal colours change together. Here, Neon.

    Themes and shortcuts, step 1 of 2: Seven themes, the same as Second Brain's, each shown in its own colours. The window changes at once and the panel stays open, so you can try them one after the other. Interface and terminal colours change together. Here, Neon.
  2. Open ◐ › Shortcuts, click an action, and press the new key.

    Every shortcut, and the voice keys (hold to talk). With Shift added, they also work inside a terminal.

    Themes and shortcuts, step 2 of 2: Every shortcut, and the voice keys (hold to talk). With Shift added, they also work inside a terminal.

More in the Docs: Shortcuts

Work4 min4 stepsReal screens

Review a diff and send comments

Read what an agent changed, file by file, comment on lines, and send your comments to it in one go.

You need
  • A session that changed files in a git repository
  1. Select the session, click View ▾ at the top and choose Review (Ctrl+Shift+R).

    The files the agent changed since its branch started, with their added and removed lines, and the diff of the first file.

    Review a diff and send comments, step 1 of 4: The files the agent changed since its branch started, with their added and removed lines, and the diff of the first file.
  2. Click a file in the list.

    Its diff, line by line, in the terminal's font. New files show whole.

    Review a diff and send comments, step 2 of 4: Its diff, line by line, in the terminal's font. New files show whole.
  3. Click a line (or Comment the file…) and write what you think of it.

    Your comment sits under the line, with the code it is about. Add as many as you like, on any file.

    Review a diff and send comments, step 3 of 4: Your comment sits under the line, with the code it is about. Add as many as you like, on any file.
  4. Click Send review.

    Every comment goes to the agent at once, each with its file, line and code, as one message in its terminal, and is marked sent to the agent. When the agent has dealt with one, mark it Handled.

    Review a diff and send comments, step 4 of 4: Every comment goes to the agent at once, each with its file, line and code, as one message in its terminal, and is marked sent to the agent. When the agent has dealt with one, mark it Handled.
If it does not work
  • Review is empty. The folder is not a git repository, or nothing changed yet. A session in its own branch (worktree) shows only its own changes.

More in the Docs: Review · Worktrees

Work4 min4 stepsReal screens

Resume, fork and find any conversation

Find an old session in the History, pick up its conversation where it stopped, or try another way on a copy.

  1. Click View ▾ at the top and choose History (Ctrl+Shift+H).

    Every session ever made, the archived ones too, with each conversation the agent had in it. Nothing is forgotten: removing a session keeps its conversations here.

    Resume, fork and find any conversation, step 1 of 4: Every session ever made, the archived ones too, with each conversation the agent had in it. Nothing is forgotten: removing a session keeps its conversations here.
  2. Type in the search box: a name, a folder, a group or a conversation id.

    The list narrows as you type. Each session shows its tool, folder and its conversations, newest first.

    Resume, fork and find any conversation, step 2 of 4: The list narrows as you type. Each session shows its tool, folder and its conversations, newest first.
  3. Click Fork… on a Claude Code session and name the copy.

    A new session starts on a copy of the conversation: the original stays as it was. Use it to try another approach without losing the first one.

    Resume, fork and find any conversation, step 3 of 4: A new session starts on a copy of the conversation: the original stays as it was. Use it to try another approach without losing the first one.
  4. To work without touching the main branch, tick In its own git branch in the New session dialog.

    The session gets a worktree beside the repository, on its own branch: its header shows the branch. Review shows only its changes.

    Resume, fork and find any conversation, step 4 of 4: The session gets a worktree beside the repository, on its own branch: its header shows the branch. Review shows only its changes.
If it does not work
  • Resume starts an empty conversation. The agent did not save that conversation (it ended before its first message). Pick another one under the session: Resume this one.

More in the Docs: History · Worktrees

Git5 min4 stepsApp only

Connect a repository (its own SSH key)

Give a GitHub or Azure DevOps repository an SSH key made and checked by Agent Manager, so sessions push without your own keys.

You need
  • A repository you can administer on GitHub, or an Azure DevOps account that reaches it
  1. Click ⎇ Repos at the top.

    The connected repositories, one per line, and a form: the repository's SSH address, and whether to use your own clone or let Agent Manager clone it. Once you have a repository, + Connect a repository opens the form.

    Connect a repository (its own SSH key), step 1 of 4: The connected repositories, one per line, and a form: the repository's SSH address, and whether to use your own clone or let Agent Manager clone it. Once you have a repository, + Connect a repository opens the form.
  2. Paste the address (git@github.com:owner/repo.git), choose your folder, and click + Add the repository.

    The form says which host it is and what the key will open. Agent Manager makes a key for this repository. For Azure DevOps the address is the one under Clone › SSH (git@ssh.dev.azure.com:v3/organization/project/repository).

    Connect a repository (its own SSH key), step 2 of 4: The form says which host it is and what the key will open. Agent Manager makes a key for this repository. For Azure DevOps the address is the one under Clone › SSH (git@ssh.dev.azure.com:v3/organization/project/repository).
  3. On GitHub, add the key (tick Allow write access if sessions should push), then click Check.

    Unfold the line (its name or its arrow): the public key with a button to copy it, the link to the repository's Deploy keys page, Check, Rotate the key, Disconnect. Once checked, the line says works and when. On Azure DevOps the key goes on your profile (User settings › SSH public keys) and opens everything your account reaches: the line says your whole account.

    Connect a repository (its own SSH key), step 3 of 4: Unfold the line (its name or its arrow): the public key with a button to copy it, the link to the repository's Deploy keys page, Check, Rotate the key, Disconnect. Once checked, the line says works and when. On Azure DevOps the key goes on your profile (User settings › SSH public keys) and opens everything your account reaches: the line says your whole account.
  4. Start a session on it: New session on its line, or in New session pick it under Repository.

    The session's git uses this key only, through GIT_SSH_COMMAND. Nothing is written in your ~/.ssh, and the agent is told how git authenticates here.

    Connect a repository (its own SSH key), step 4 of 4: The session's git uses this key only, through GIT_SSH_COMMAND. Nothing is written in your ~/.ssh, and the agent is told how git authenticates here.
If it does not work
  • Check says write is refused. On GitHub, the deploy key needs Allow write access.
  • On Azure DevOps the key does not work. It must be on your profile (User settings › SSH public keys), SSH must be allowed for the organization, and the key must not have expired: the line says which.
  • ssh -T git@github.com shows another account. That command ignores the session's key. Check with git ls-remote origin instead.

More in the Docs: Repositories and their keys

Voice4 min4 stepsApp only

Talk to your agents

Dictate with F8 into a box you send with Enter, and drive the app with F9: approve, go to a session, ask who waits.

You need
  • An OpenAI key in ◐ › Keys, for the transcription (Keys)
  1. Open ◐ › Keys and paste your OpenAI key.

    On Windows it is encrypted for your Windows user (DPAPI); on Linux, the file is readable by you only. It is never shown again, never written to a log. Your Second Brain keys are listed below it.

    Talk to your agents, step 1 of 4: On Windows it is encrypted for your Windows user (DPAPI); on Linux, the file is readable by you only. It is never shown again, never written to a log. Your Second Brain keys are listed below it.
  2. In a session, hold F8 and speak. Release it when you are done.

    A small box at the bottom shows what you say, phrase by phrase. On release the text stays in the box: read it, correct it, hold F8 again to add more, then press Enter: all of it goes to the session. Nothing is sent before your Enter.

    Talk to your agents, step 2 of 4: A small box at the bottom shows what you say, phrase by phrase. On release the text stays in the box: read it, correct it, hold F8 again to add more, then press Enter: all of it goes to the session. Nothing is sent before your Enter.
  3. Hold F9 and give a command: approve Flaky tests.

    Agent Manager asks before acting, with what it heard: « Approve “Flaky tests”? ». Other commands: go to …, grid view, who is waiting, wrap up …, new session ….

    Talk to your agents, step 3 of 4: Agent Manager asks before acting, with what it heard: « Approve “Flaky tests”? ». Other commands: go to …, grid view, who is waiting, wrap up …, new session ….
  4. Click OK.

    The OK goes to that agent, and it carries on. A soft sound tells you when an agent waits for you; ♪ at the top right mutes it.

    Talk to your agents, step 4 of 4: The OK goes to that agent, and it carries on. A soft sound tells you when an agent waits for you; ♪ at the top right mutes it.
If it does not work
  • Nothing is written after F8. Check the key in ◐ › Keys, and that the microphone is allowed for Agent Manager. The toast says what went wrong.
  • F9 put my words in the box. When what you say is not a command, it is dictation: it joins the text of the box, and Enter sends it.
  • I closed the box. Its text is kept. Ctrl+K › Dictation box opens it again; Earlier brings back what you sent or emptied.

More in the Docs: Voice · Keys

Agents4 min3 stepsReal screens

Let an agent start helpers (MCP)

Have a lead agent split a job and start helper sessions of its own, which you can watch and take over.

You need
  • Claude Code or Codex (each session is offered Agent Manager's MCP server)
  1. Ask a lead agent for a job that splits well: prepare the release: run the tests in one helper and draft the changelog in another.

    The agent sees Agent Manager's tools: list the sessions, read a screen, start a session, send to its own children, wait for one.

    Let an agent start helpers (MCP), step 1 of 3: The agent sees Agent Manager's tools: list the sessions, read a screen, start a session, send to its own children, wait for one.
  2. Watch the list.

    Its helpers appear under it, marked ↳ Release lead, in real terminals. Open any of them: you can type in it, stop it, take over.

    Let an agent start helpers (MCP), step 2 of 3: Its helpers appear under it, marked ↳ Release lead, in real terminals. Open any of them: you can type in it, stop it, take over.
  3. Open the Grid view on the lead and its helpers.

    The lead reads its helpers' screens and waits for them; it never types into a session that is not its own, and never approves anything.

    Let an agent start helpers (MCP), step 3 of 3: The lead reads its helpers' screens and waits for them; it never types into a session that is not its own, and never approves anything.
If it does not work
  • The agent says it has no such tool. The session was started before the MCP server was offered: Restart it.
  • You want the same tools outside the app. ◐ › Connect other tools gives the command for Claude Code, Codex or any MCP client.

More in the Docs: The MCP server

Agents2 min2 stepsReal screens

Follow a session from your phone

Keep an eye on a Claude Code session, and answer it, from claude.ai/code or the Claude app while you are away.

You need
  • Claude Code signed in with a Claude account that has Remote Control
  1. In New session, keep Remote Control: also on claude.ai/code and the Claude app ticked (it is on by default for Claude Code).

    The session is started with Claude Code's own Remote Control, under the name Group · Session, so you recognise it on the other side.

    Follow a session from your phone, step 1 of 2: The session is started with Claude Code's own Remote Control, under the name Group · Session, so you recognise it on the other side.
  2. Start it, then open claude.ai/code or the Claude app on your phone.

    Its card shows ⇄ claude.ai. The same conversation is there, live: read it, answer it. Back at your desk, the terminal shows what happened.

    Follow a session from your phone, step 2 of 2: Its card shows ⇄ claude.ai. The same conversation is there, live: read it, answer it. Back at your desk, the terminal shows what happened.
If it does not work
  • The session is not on claude.ai. Remote Control is Claude Code's own feature and needs your Claude account to have it. Untick it for sessions that should stay on this computer only.

More in the Docs: Remote Control

Second Brain4 min4 stepsReal screens

A session that documents itself in the Brain

Start an agent that reads where the project stands in your Second Brain, then proposes your words, its decisions and a journal as it works.

You need
  • Second Brain GD running on the same computer: the status bar says ✦ Second Brain connected
  1. In New session, keep Share the Second Brain ticked and tick ↳ Document this session in the Brain as it works.

    The session gets a Brain key of its own, at the « propose » level: it can read and search, and what it proposes waits for the Brain's archivist.

    A session that documents itself in the Brain, step 1 of 4: The session gets a Brain key of its own, at the « propose » level: it can read and search, and what it proposes waits for the Brain's archivist.
  2. Start it.

    Its card shows ✦ Brain and its header shares the Second Brain. Before changing anything, the agent reads the project card, the last journal and the open questions, and checks them against git log.

    A session that documents itself in the Brain, step 2 of 4: Its card shows ✦ Brain and its header shares the Second Brain. Before changing anything, the agent reads the project card, the last journal and the open questions, and checks them against git log.
  3. Open ◐ › Keys.

    Under Second Brain keys made by Agent Manager, one key per session, with its calls. A key nothing uses any more can be revoked; removing a session revokes its key.

    A session that documents itself in the Brain, step 3 of 4: Under Second Brain keys made by Agent Manager, one key per session, with its calls. A key nothing uses any more can be revoked; removing a session revokes its key.
  4. In your Second Brain, open AI tools and your Brain.

    What sessions proposed: your directives, their decisions, a journal at each commit, each with who proposed it and why. Approve, reject, or let the archivist file them with your rules.

    A session that documents itself in the Brain, step 4 of 4: What sessions proposed: your directives, their decisions, a journal at each commit, each with who proposed it and why. Approve, reject, or let the archivist file them with your rules.
If it does not work
  • The Brain options are greyed out. Second Brain is not running, or not found: the status bar says why.
  • Document is greyed out. Claude Code only for now, on this computer.

More in the Docs: Second Brain

Cloud6 min2 stepsApp only

Run an agent in a cloud sandbox

Give an agent a throw-away machine on Fly.io, see what it costs as it runs, and keep a budget that only pauses.

You need
  • A Fly.io account and its API token in ◐ › Keys
  • An Anthropic API key (Claude Code) or OpenAI key (Codex) in Keys, for the agent in the cloud
  1. Open ☁ Cloud.

    Three steps: the image (what it holds: Ubuntu, tmux, git, Claude Code, Codex), deploy with an estimate first, and this month's costs against your budget. Below, a live card per sandbox: the machine, the SSH tunnel and its latency, the branch, files not committed, commits not pushed, and what it cost today.

    Run an agent in a cloud sandbox, step 1 of 2: Three steps: the image (what it holds: Ubuntu, tmux, git, Claude Code, Codex), deploy with an estimate first, and this month's costs against your budget. Below, a live card per sandbox: the machine, the SSH tunnel and its latency, the branch, files not committed, commits not pushed, and what it cost today.
  2. Click A machine + an agentic coding session, or pick ☁ A cloud sandbox under Where in New session.

    The same dialog as for this computer. The machine starts in about 30 seconds; the session's terminal is the agent's terminal in the cloud, through an SSH tunnel. It reaches your Second Brain through that tunnel, with a key of its own.

    Run an agent in a cloud sandbox, step 2 of 2: The same dialog as for this computer. The machine starts in about 30 seconds; the session's terminal is the agent's terminal in the cloud, through an SSH tunnel. It reaches your Second Brain through that tunnel, with a key of its own.
If it does not work
  • Deploy is refused. The month's budget is reached: raise it, or wait. Budgets and caps only pause machines, never destroy them.
  • Costs differ from Fly's invoice. Agent Manager measures running time from Fly's machine history × the published price. The invoice stays on Fly's dashboard.

More in the Docs: Cloud sandboxes · Keys