MATTERAI / ORBCODE CLINEW

AGENTIC CODING
IN YOUR
TERMINAL.

A standalone terminal coding agent powered by Axon models. Streaming TUI, live reasoning, edit/command approvals, headless mode, and Claude-Code-compatible hooks.

$ npm install -g @matterailab/orbcode
VIEW DOCS
orbcode — zsh
refactor the auth module to use async/await
✦ Thinking…
▸ Read File
src/auth/auth.ts
▸ Search Files
"login" in src/**/*.ts
▸ Multi File Edit
3 files · Added 24 lines, removed 18 lines
✓ Execute Command
npm run typecheck
Refactored src/auth/auth.ts and two callers to use async/await. The callback-based login() is now an async function, and verifyToken() returns a Promise. Typecheck passes.
commit it
0k
Context window
0
Max tool steps per turn
0
Built-in Axon models
CAPABILITIES

BUILT FOR
THE TERMINAL.

Streaming TUI

Live Markdown

Streaming responses rendered as markdown with a lightweight ANSI renderer. Headers, lists, code fences, inline code, and links.

Reasoning

Live Thinking

Reasoning streams under a collapsible ✦ Thinking block and collapses to ✦ Thought for Ns when done. Ctrl+O toggles the whole transcript.

Tool Rows

Transparent Actions

Each tool call shows a formatted name, a one-line summary, live running state, then ✓/✗ with a short result preview.

Edit Diffs

Real Diffs

File-modifying tools render a real diff with a stats header and red/green backgrounds, both in the approval prompt and the finished tool row.

Approvals

Safe by Default

Read-only tools run freely. File edits and commands prompt first. Dangerous commands can never be auto-approved.

Headless Mode

Scriptable

Pipe a single prompt with -p for scripts and CI. Prints only the final content to stdout; errors go to stderr.

Hooks

Claude-Code Compatible

Block, auto-approve, rewrite, or inject context at fixed points in the agent loop. Scripts written for Claude Code work here.

@-References

Fuzzy File Search

Type @ in the input to fuzzy-search workspace files. ↑/↓ to choose, enter/tab inserts the match into the prompt.

Tasks

Todo Tracking

The model maintains a checklist via update_todo_list, rendered as a compact Tasks panel with pending, in-progress, and done states.

MODEL FAMILY

THREE AXON MODELS,
BUILT IN.

The same Axon models that power Orbital, available in your terminal. Switch with /model, or declare your own via customModels in settings.json.

$3/M in · $9/M out

axon-eido-3-code-pro

Frontier model for the hardest agentic tasks.

  • 400k context, 64k output
  • Maximum intelligence and reasoning
  • Best for complex, long-running tasks
$2/M in · $6/M out

axon-code-2-5-pro

High-intelligence model for all-day coding. The default.

  • 400k context, 64k output
  • Complex architectural decisions
  • Best for professional development
free

axon-code-2-5-mini

Lightweight and fast model for everyday tasks.

  • 400k context, 64k output
  • Low-effort tasks and quick fixes
  • Cost-effective for high-volume usage
SAFETY

APPROVALS &
CONTROL.

Read-only tools run without prompting. File edits and commands prompt first, with the exact target or command line shown before anything runs. Dangerous commands — deletes, force-pushes, system changes — can never be auto-approved.

Ask

Prompt for every edit and command. The default.

Accept Edits

Auto-approve file edits; commands still prompt.

Auto-Approve

Auto-approve safe commands too. Dangerous ones still prompt.

Deny Feedback

Denials are reported back so the model adjusts course.

Cycle modes at runtime with Shift+Tab.

approval prompt
▸ Execute Command
git push --force origin main
⚠ dangerous: force-push
Allow this command?
[y] allow once[n] deny[a] allow session
dangerous commands have no auto-approve option
BEYOND INTERACTIVE

SCRIPTABLE &
EXTENSIBLE.

HEADLESS MODE

Pipe a single prompt

Run OrbCode non-interactively with -p. Prints only the final content to stdout — perfect for scripts and CI. Add --yolo to auto-approve edits.

$ orbcode -p "fix the lint errors" --yolo
HOOKS

Claude-Code compatible

Shell commands run at fixed points in the agent loop. Block dangerous actions, auto-approve trusted ones, rewrite tool inputs, or inject context. Scripts written for Claude Code work here.

"PreToolUse": [{ "matcher": "execute_command" ... }]

START CODING
IN THE TERMINAL.

Install globally with npm, sign in once, and run orbcode from any project directory. Open source under the MIT license.

$ npm install -g @matterailab/orbcode
VIEW ON GITHUB