TotalApp Docs

Dev Hub Terminal

AI-assisted developer terminal for code analysis, git commands, and GitHub integration.

Overview

Dev Hub is a multi-block terminal designed for developers who want structured, AI-enriched output alongside standard git and version control operations. Unlike a plain shell, Dev Hub renders each command result as a typed block — RunBlocks show step-by-step progress, DiffBlocks show before/after code diffs, and PipeBlocks track pipeline stage completion.

Compact mode is available to reduce the vertical footprint of each block — useful when you want to keep multiple command results visible at once. Toggle it with the compact icon in the toolbar.

Dev Hub Terminal compact
RunBlock done
dev$ git log --oneline
→ Fetching commit history...
eda28b7 add AI panel close tab
8e771ab fix index.css selector
d0072a3 improve kanban board
DiffBlock — refactor-file src/utils/helpers.ts done
BEFORE
function calc(a,b) {
  return a + b;
}
AFTER
const calc = (
  a: number, b: number
) => a + b;
PipeBlock — 2 stages running
git log --oneline
analyze-performance src/index.ts
dev$

Git Commands

Run standard git commands directly in Dev Hub. Output is rendered as a RunBlock with progress steps.

CommandDescription
git logShow recent commit history with hashes, authors, and messages
git log --onelineCompact one-line-per-commit history view
git log --oneline --graphASCII branch graph alongside commit history
git statusWorking tree status — staged, unstaged, and untracked files
git diffShow unstaged changes as a diff
git blame [file]Line-by-line author attribution for a file
dev$ git log --oneline
→ Fetching commit history...
eda28b7 add ChevronRight close tab to left edge of AI panels
8e771ab fix index.css selector
d0072a3 improve kanban board

GitHub / GitLab / Bitbucket Commands

Dev Hub can query your connected version control platforms. Authenticate in Settings → General by providing your access tokens. Commands fetch live data and render it in structured RunBlocks.

CommandDescription
list-reposList all repositories for the authenticated account
get-issues [repo]Fetch open issues for the specified repo
pr-list [repo]List open pull requests with titles, authors, and status
get-commits [repo]Recent commits for a specific repository

Authentication Required

GitHub, GitLab, and Bitbucket commands require access tokens. Go to Settings → General to enter your tokens. Without a token, these commands return an authentication error.

AI Commands

Dev Hub integrates AI analysis directly into the terminal session. AI commands send file contents or git output to Claude and render the result as a DiffBlock (for refactoring) or a TextBlock (for analysis and explanation).

CommandBlock TypeDescription
analyze-performance [file]TextBlockAI identifies performance bottlenecks, inefficient loops, and memory issues in the file
explain-function [file]TextBlockAI explains what each function does, its parameters, return values, and side effects
refactor-file [file]DiffBlockAI suggests refactored code — shown as a before/after diff with explanatory comments
dev$ refactor-file src/utils/helpers.ts
→ Reading file...
→ Analysing structure...
→ Generating refactored version...
[DiffBlock rendered with before/after code]

Pipeline Mode

Chain commands together using the pipe operator |. Each stage in the pipeline is tracked in a PipeBlock, with status indicators showing which stages have completed and which are running.

dev$ git log --oneline | analyze-performance src/index.ts
→ [Stage 1] git log --oneline ... done
→ [Stage 2] analyze-performance ... running

Pipelines short-circuit on error — if any stage fails, subsequent stages are skipped and the PipeBlock shows the failed stage in red.

Example Session

Here is a typical Dev Hub session for investigating a performance issue and drafting a fix:

  1. Run git log --oneline --graph to see recent commit history and identify when the issue was introduced.
  2. Run git blame src/components/HeavyComponent.tsx to find which author last modified the suspect code.
  3. Run analyze-performance src/components/HeavyComponent.tsx to get an AI summary of performance problems.
  4. Run refactor-file src/components/HeavyComponent.tsx to see the AI's suggested rewrite as a DiffBlock.
  5. Copy the refactored code from the DiffBlock into your editor to apply the changes.

AI Assistant

Open the built-in AI Assistant from the tab on the right edge of the screen for help deciding what to run. It knows the Dev Hub service catalog (Performance Analysis, AI Refactoring, Function Explainer) and your current repository/selection, so you can ask things like:

  • "Which service should I run for a suspected memory leak?"
  • "What exactly does AI Refactoring check for?"
  • "What are best practices for refactoring legacy code safely?"
  • "What should I do after running Performance Analysis?"

The assistant does not see terminal output — it only knows the service catalog and your current repo/selection, so it recommends and explains rather than interpreting results you've already run.