> ## Documentation Index
> Fetch the complete documentation index at: https://docs.alpha.new/llms.txt
> Use this file to discover all available pages before exploring further.
# Follow the agent's work
> Read what the agent did in plain English, open any command or edit, follow helpers and workflows, and see what went wrong when a turn fails.
The transcript is the running record of a session: your messages, the agent's replies, and every command, edit and thought in between, folded into short lines you can open. Turns run on Alpha's servers, not in your tab, so the transcript is complete when you come back. To change what the agent is doing mid-turn, see [Queue, steer or stop](https://docs.alpha.new/chat/queue-and-steer.md); for plans and the agent's questions, see [Plan mode](https://docs.alpha.new/chat/plan-mode.md).
## What does a turn look like?
A turn runs from your message until the agent stops. Everything the agent does between two replies collapses into one line, so even a long turn stays easy to read.
| You see | What it is |
| - | - |
| Your message, in a tinted bubble on the right | What you sent, with any attached context as chips. A message sent mid-turn carries "Steered into the running turn". |
| A folded row: icons, a tally, a chevron | Every action between two replies, such as "2 tools · 8 commands · 7 thoughts". |
| The reply | Markdown with code blocks, tables and links. Some replies are [interactive](https://docs.alpha.new/chat/interactive-answers.md). |
| Cards | Plans, questions, sub-agents, workflows and errors. Cards are never folded. |
| **Copy** | Under the last reply, and under any other message, yours included, on hover. Hidden while the agent works. |
## Open the folded actions
The tally counts each kind of action, in the order it first appeared and with thoughts last. The kinds are **command**, **edit**, **search**, **plan update**, **tool** and **thought**. While an action runs, its icon bobs and the tally shimmers. In a narrow column, hover the row to read the whole tally.
Click the row and each action becomes one plain-English line, such as "Ran the tests", "Started the dev server", "Edited BookingForm.tsx" or "Queried the database". A running line reads in the present ("Building the project"). A failed one ends in a red pill such as **exit 1**. A failed command is not a failed turn: the agent usually reads the error and fixes it on the next line.
Click a line to open its detail:
| Action | Its detail |
| - | - |
| Command | The command, then its output under **output · N lines**. A command of several lines shows in full, with **collapse** to fold it to its first line. |
| Edit | Each file touched, marked **+** created, **−** deleted or **\~** changed. Type `/diff` for the full [diff](https://docs.alpha.new/reference/slash-commands.md). |
| Web search | The queries, or the page it opened, and the results as links. |
| Thought | The reasoning itself. The line shows the model's title for it, its first sentence, or "Thought it through". |
A line that reads a [skill](https://docs.alpha.new/skills/overview.md), such as "Read the slides skill", opens nothing. Its detail stays closed.
Everything the agent did between two replies, folded into one line with a tally.
## What does the agent do while I wait?
Between outputs, a pixel loader shows a verb that changes every 3 seconds ("Forging…", "Thinking…", "Reading the workspace…") and a timer. The timer appears after 3 seconds and restarts with each pause, so it measures this gap, not the whole turn. The verb gets specific when there is something specific to say:
| It reads | Meaning |
| - | - |
| A status note, such as "Loading the installed MCP servers into this session…" | Alpha is getting the session's machine ready for the agent. |
| "Waiting for your answer" | A [question card](https://docs.alpha.new/chat/plan-mode.md#answer-the-agents-questions) is open, and the turn waits for you. The timer is hidden. |
| "Waiting for your approval" | An older Claude Code session on **Ask me** wants [an action approved](https://docs.alpha.new/chat/plan-mode.md#approve-actions-in-older-claude-code-sessions). |
| "Sending your answer" | Your answer is on its way to the agent. |
| "Robie is working", "3 agents are working" | Sub-agents are doing the work. |
| "Stopping" | You pressed Stop, and the turn is ending. |
Between outputs, the loader shows a verb and how long this pause has lasted.
## Open files from the chat
A file the agent names in its reply is a link. Clicking it opens the file as a [workspace](https://docs.alpha.new/workspace/files.md) tab, at the line when one is given. Hover to see where it goes ("Open src/app.ts:42").
| The agent writes | Clicking it |
| - | - |
| `[app.py](/workspace/app.py)` or `file:///workspace/README.md` | Opens the file in a tab. |
| `[src/app.ts:42](src/app.ts:42)` or `#L42` | Opens the file at line 42 and marks it. |
| `` | Shows the image in the reply; clicking opens the [image viewer](https://docs.alpha.new/workspace/file-viewers.md). |
| A `/s/…` link | Shows as a session badge, and opens that session in a new tab. |
| A web address | Opens in a new browser tab. |
Paths outside `/workspace`, folders and addresses such as `localhost:3000` aren't opened as files.
## Follow sub-agents and workflows
The agent can split a task between helpers, called sub-agents, each with its own task, working in the same machine.
### Sub-agent cards
Each sub-agent gets a card where it was spawned. The header shows its name (Codex gives each one a nickname), its role, its model and effort, and a status pill with its time. The body shows its task (**more** for the rest), a live feed of its steps as action lines, its to-do list, and its answer once done.
| Pill | Meaning |
| - | - |
| **starting**, **working** | The sub-agent is at work. |
| **done** | It finished and handed back its answer. |
| **failed** | It hit an error, shown in its card. |
| **interrupted** | It was interrupted before it finished. |
| **closed** | The main agent closed it before it answered. |
| **stopped** | The turn ended while it was still working. |
A roster above the composer lists the turn's sub-agents: a dot each, up to three names with **+N**, and a count such as **2 working** (or **3 done** once they finish). Click it to see each agent's activity; click an agent ("Show this agent in the transcript") to jump to its card.
Sub-agents never stop to ask you anything. Alpha tells them no one is there to answer, and they carry on with their own judgement, so every question card you see comes from the main agent. In Claude Code, the turn keeps running until its last sub-agent finishes.
### Workflow cards
Claude Code only A workflow is a team of agents Claude runs in the background, in phases. With the session's effort on **ultracode**, Claude runs one for every substantive task, and it may start one in any session. Writing the word "ultracode" in a message asks for one for that turn; see [Choose the agent and model](https://docs.alpha.new/chat/agents-and-models.md#when-is-ultracode-worth-it). When the workflow finishes, Claude follows up on its result, and your turn keeps running until then.
The card's pill counts finished agents, such as **working · 3/8**, then reads **done**, **failed** or **stopped**. The card lists each phase, with each agent's current tool, tokens, time and result. **Stop** on the card ("Stop this workflow and its agents") stops that workflow only; the composer's Stop ends the whole turn, workflows included.
> [!WARNING]
>
> A workflow runs many agents at once, so it costs several times the tokens and time of an ordinary turn. Stop one you don't need from its card.
> [!NOTE]
>
> If you don't see Claude Code on the home screen, it isn't turned on for your Alpha.
## Find, copy and export
A long session opens with its newest 2 turns, so it appears at once. Older turns load as you scroll up, or all at once from the "Show N earlier messages" marker at the top. Press ⌘F to find text: Alpha first shows every loaded message so the browser can search them. Turns not loaded yet aren't searched until you scroll to them. Use Ctrl instead of ⌘ on Windows and Linux.
To export the transcript, scroll to the top of the session, hover the first message and click **More** (⋯) → **Export as Markdown**.
The transcript is append-only: you can't edit a sent message or regenerate a reply. To try another direction and keep this one, [fork the session](https://docs.alpha.new/sessions/fork-and-delete.md). Values of your [secrets](https://docs.alpha.new/secrets.md) that are 8 or more characters long show as `[redacted]`, in the transcript and in exports. An email address or a plain web address is left as it is.
## What if a turn fails?
A failed turn shows one red card saying what failed, sometimes with a **details** fold. The workspace and the transcript up to the failure are kept, and your next message carries on from there.
| You see | What to do |
| - | - |
| The model provider's own message, or "model provider error" when it gave none | Send again. If it repeats, try [another model](https://docs.alpha.new/chat/agents-and-models.md#choose-a-model). |
| "The agent exited with code N" | Open **details** to read why, then send again. |
| "Couldn't pick the turn back up: …" | Alpha's server restarted during the turn and couldn't reconnect to it. Ask the agent to carry on. |
| "Claude sign-in: this session has no Claude credentials. …" | Open **Account menu → Sign in with ChatGPT/Claude** and sign in on the **Claude** row ([how](https://docs.alpha.new/account/usage-and-costs.md#run-claude-code-sessions-on-your-claude-plan)), then resend. |
| "Claude sign-in: your Claude subscription token was rejected (…)" | Sign in with Claude again, then resend. |
| "Claude sign-in: the server's Anthropic key was rejected (…)" | Ask whoever runs your Alpha to fix the key. |
| "The session's machine cannot be reached right now. Nothing was discarded — …" | Send again in a moment. If your message was refused, its text goes back into the composer. Resetting is a [last resort](https://docs.alpha.new/sessions/machines.md#reset-the-sessions-machine). |
A turn you stop isn't a failure. For other messages, see [Troubleshooting and FAQ](https://docs.alpha.new/reference/troubleshooting.md).
## Next steps
> **[Queue, steer or stop](https://docs.alpha.new/chat/queue-and-steer.md)**
>
> Send a second message without waiting, or stop a turn heading the wrong way.
> **[Plan mode and the agent's questions](https://docs.alpha.new/chat/plan-mode.md)**
>
> Answer the cards that make a turn wait, and approve a plan before any code.
> **[Choose the agent and model](https://docs.alpha.new/chat/agents-and-models.md)**
>
> Pick Codex or Claude Code, set the effort, and turn on ultracode workflows.
> **[Files and uploads](https://docs.alpha.new/workspace/files.md)**
>
> Where the files the agent links to live, and how to download them.
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.