🦀 v0.16.4 · beta · MIT licensed

Gajae Code

Encode intention. Decode software.

An external coding-agent harness for interviews, approved plans, and durable verification. Run gjc from the repository or worktree you choose, on the coding plan you already pay for — then connect bots, editors, and agent shells through the SDK session CLI, Coordinator MCP, or ACP.

shape · interview act · reviewed plans prove · evidence
— · stars — · downloads/mo
Download latest Read the Docs
Gajae Code autonomous coding-agent hero illustration
What is Gajae Code?

A focused harness, not a hidden plugin

Gajae Code (gjc) runs from the repository or worktree you choose, then gives the agent a deliberately small workflow: clarify before guessing, plan before mutation, and track every goal to completion with evidence.

It runs beside your existing tools — start it alongside Codex CLI, Claude Code, OpenCode, or Claw Code when you want structured planning, persistent evidence, an SDK integration, or an isolated worktree.

recommended workflow 
# run gjc in the repository or worktree you want to change
$ gjc


# inside the session: shape -> act -> prove
> /skill:deep-interview # clarify intent
> /skill:ralplan        # plan & critique
> gjc ultragoal create-goals --brief-file plan.md
> gjc ultragoal complete-goals
# optional: /skill:autoresearch before the interview or after the spec
Latest stable · v0.16.4

Gajae Code v0.16.4

Published 2026-09-05. Release binaries and npm packages are available.

Read what’s new GitHub Release
Providers

Bring your coding plan

Log in once and run GJC on the subscription you already pay for. Run /login inside a session and pick your plan — no second bill for the same tokens.

OAuth plans

Plan / subscriptionOAuth login
Claude Pro / Maxanthropic
ChatGPT Plus / Pro (Codex)openai-codex (browser) · openai-codex-device (headless)
Cursorcursor
GitHub Copilotgithub-copilot
OpenCode Zen / OpenCode Goopencode-zen · opencode-go
Kimi Code / Coding Plan / Moonshotkimi-code · moonshot
Z.AI GLM Coding Planzai
MiniMax Coding Plan (Intl / CN)minimax-code · minimax-code-cn
xAI (Grok)xai
Alibaba Token Plan / Qwen Portalalibaba-token-plan · qwen-portal

More OAuth plans — Google Gemini CLI, GitLab Duo, Perplexity Pro/Max, Fire Pass, Xiaomi Token Plan — are covered in the models reference.

Coding-plan presets

Key-based plans onboard with one command. A preset writes the API type, base URL, env var, compatibility flags, and a live model catalog together, so new models appear without a GJC update.

gjc setup provider --preset commandcode-goat # Command Code GOAT (CMD_API_KEY) gjc setup provider --preset cline-pass # ClinePass (CLINE_API_KEY)
  • Command Code GOAT — pulls the provider’s live /models catalog; claude-* models route through native Anthropic Messages, everything else through Chat Completions. Aliases: commandcode, goat.
  • ClinePass — no hardcoded models; GJC fetches Cline’s live catalog the same way Cline does. Aliases: clinepass, cline.
  • Also available: minimax, minimax-cn, glm, alibaba-token-plan — or /provider add --preset <name> inside the TUI.

Beyond coding plans: 50+ API-key providers, local runtimes (Ollama, LM Studio, vLLM), and gateways (Cloudflare AI Gateway, Vercel AI Gateway, LiteLLM) all work. Register endpoints in models.yml, pool multiple accounts per provider with usage-aware routing, mix vendors per agent role with presets and profiles, or centralize team credentials with the auth broker/gateway.

External connections

Choose the client boundary that fits

Drive one session over the SDK, fan work out across worktrees with Coordinator MCP, or run GJC inside an agent shell over ACP. Every route is credential-free JSON — never terminal scraping. Legacy Bridge and RPC modes are not current interfaces.

SDK & Migration

Build direct session clients on the loopback SDK WebSocket. The preserved route also explains migration away from historical Bridge and RPC ingress modes.

Bots & controllers

OpenClaw, Hermes, Grokbot, a Discord bot, or a cron script drives real sessions through the broker-bound SDK session CLI (gjc sdk session list · send · status · tail) and the bundled gjc-sdk-discover / operate / author procedures.

Coordinator MCP

Need event-driven fan-out across many worktrees instead of one session at a time? gjc mcp-serve coordinator — installed by gjc setup hermes — exposes the delegation tools for that shape.

Run GJC inside an agent shell

HostSupportWhat you get
Paseo★★★★★
first-class		Native ACP provider installed by GJC itself: model catalog, Default/Plan modes, thinking levels, real permission prompts, cancel that terminates owned subagents, mobile control.
Orca★★★★☆
works, one field		GJC runs as a custom CLI agent, one worktree per session, with diff review, terminal splits, SSH worktrees, and the mobile companion. No usage tracking or account hot-swap yet.
T3 Code★★★☆☆
experimental		No GJC harness upstream yet — T3 Code ships harnesses for Codex, Claude, Cursor, Grok, and OpenCode only. Run GJC beside it; the native provider is proposed upstream.

Paseo, in one paste

gjc setup paseo # writes the ACP provider entry, backs up, never restarts your daemon paseo daemon restart paseo provider ls # gjc must read `available` paseo run --provider gjc --cwd /path/to/repo "your prompt" gjc setup paseo --check # pass / stale / drift, with a machine-readable --json gjc setup paseo --remove # rolls back only the keys GJC itself created

Orca, in one field: install GJC, then add a custom agent with command gjc and no arguments. Orca pre-fills a permission-bypass flag for agents that expose one — GJC has none by design, so leave the arguments empty and keep GJC’s own approval gates.

Long prompts are safe to leave running: the SDK prompt deadline is a progress-aware inactivity lease (sdk.promptDeadlineMs, 30 min default) bounded by sdk.promptMaxRuntimeMs (6 h default), renewed only by attributable tool execution for the accepted turn.

The method

Shape · Act · Prove

One small loop, done well. No sprawling skill zoo — Gajae Code improves by making this method better.

01 — Shape

Interview before guessing

deep-interview turns vague requests into concrete requirements, boundaries, and explicit non-goals before any code is touched.

02 — Act

Plan before mutation

ralplan builds and critiques the implementation plan, then ultragoal carries the approved plan through execution, revision, verification, and evidence.

03 — Prove

Finish with evidence

ultragoal tracks goals, revisions, checks, and completion evidence so a finished task ships with a verifiable trail, not a hopeful claim.

Workflow surface

A path you can read

Clarify, plan, and prove on a three-step loop. Autoresearch is the fourth skill: an optional evidence mission before an interview or after a spec — not the next arrow after ultragoal.

deep-interview

Clarify ambiguous requirements before planning or code changes.

ralplan

Build and critique an implementation plan before mutation.

ultragoal

Track goals through execution, revision, verification, and evidence.

Optional: /skill:autoresearch when evidence must ground that loop — web, REPL/data, or both — then a structured verdict, never product code.

Capabilities

Built for reviewable autonomy

Everything you need to run an agent you can actually trust with a repository.

Interview-first intent

Requirements are clarified up front with deep-interview, so the agent works from a shared, concrete understanding instead of a guess.

Reviewed plans

ralplan drafts and critiques the approach through planner, architect, and critic perspectives before a single file is mutated.

Durable evidence

ultragoal records goals, revisions, checks, and completion evidence under .gjc/ — a verifiable trail for every task.

Client integrations

Use the SDK WebSocket to control a direct session, Coordinator MCP to automate multiple sessions, or ACP to connect an editor client.

Worktree isolation

Create an isolated git worktree from a new branch for risky or reviewable work, keeping the primary checkout clean.

Coordinator MCP

Coordinate multiple sessions with bounded turns, structured questions, durable status, and reports through the Coordinator MCP surface.

Autoresearch

Run /skill:autoresearch for a goal-directed research mission that interleaves web research with the persistent Python REPL and ends on a structured verdict — never product code.

Your own coding plan

Run /login and use the subscription you already pay for — Claude, Codex, Cursor, Copilot, OpenCode, Kimi, GLM, MiniMax, Grok, Qwen — or onboard a key-based plan with gjc setup provider --preset.
Architecture

A layered runtime

From CLI launch to model streams, every layer has a clear responsibility.

CLI & client surfacescoding-agent
The gjc binary registers commands (setup, deep-interview, ralplan, ultragoal, autoresearch) and supports interactive sessions, print output, SDK WebSocket clients, Coordinator MCP automation, and ACP editor clients.
Agent runtimeagent
A stateful turn loop transforms context, calls the model stream, executes tool calls, appends results, and emits lifecycle events.
Provider boundaryai
A model registry, provider implementations, auth, streaming, usage accounting, and retry/overflow handling sit behind one clean interface.
Toolsread · edit · bash · lsp
A built-in tool registry for file, code, and runtime operations — reading, editing, AST tools, search, LSP, the browser, subagents, and more.
Native & TUInatives · tui
Rust N-API helpers accelerate search, diff, syntax, and shell work; the terminal UI framework renders the interactive experience.

Explore the architecture →

Get started

Install the latest GJC binary

No Bun required. We detect your platform and link directly to the newest standalone release.

# 1 · choose your platform
View latest downloads

Choose the platform and architecture for this machine.

Need another build? View every latest release asset. Read the full installation guide. Standalone binary source availability remains on GitHub.

Workflow surface

Four skills, four role agents

Four skills, four role agents — a surface people can hold without overhead. Parallel workers coordinate over IRC (in-session agent DM/broadcast). Multi-session machines still use Coordinator MCP or the SDK — not a fifth skill.

Workflow skills

SkillWhat it does
deep-interviewClarifies ambiguous requirements before planning or code changes.
ralplanBuilds and critiques an implementation plan before mutation.
ultragoalTracks goals through execution, revision, verification, and evidence.
autoresearchRuns goal-directed research missions (web + REPL/data) and ends on a structured verdict.

Role agents

AgentWhat it does
executorBounded implementation, fixes, and refactors.
architectRead-only architecture and code-review assessment.
plannerRead-only sequencing and acceptance criteria.
criticRead-only plan critique and actionability review.

Skills reference →

Mascots

Meet the Gajae pets

GJC ships three terminal pets — pick one with /pet and it lives on your status line, rendered over Kitty, Sixel, and iTerm2. These are the exact 16×16 pixel frames, palettes and step timings from the source: play their idle loop, their work animation, or their signature pose, and click a card to apply its paired theme.

RedGajae

The Red Crab, who likes to work-out.

The default skin. It idles through glances and visor flickers, and throws both claws up in a ^^ flex when it shows off.

BlueGajae

The Blue Crab, who wants to rest.

The same crab in blue-crab colors — crabShell shell, claw highlight, azure belly. Its signature burst is a little >< sob with foam tears.

Ouroboros

The little snake who keeps going.

Blinks, flicks its tongue upward, occasionally sobs. Agent work enters an eight-frame infinity loop, and its heart flex rolls the coil into a 180° heart pose.

Pet pixel art by @project-catsy, the artist behind RedGajae, BlueGajae, and Ouroboros.

Appearance

Two crabs and a serpent

GJC compiles its bundled themes into the binary — browse them live with /theme. Every swatch below is copied verbatim from the shipped theme files.

red-claw

default dark

🦀 brand · 🦞 agents · 🦀 running · ▸ bullets

Warm ink surfaces under claw-orange and crab-shell red — the TUI’s default identity.

blue-crab

light slot

🦀 brand · 🦞 agents · 🌊 running · ▸ bullets

Deep-ocean navy with crabShell blues and foam text — the light-appearance default.

ouroboros

dark

🐍 brand · ∞ running · ◇ bullets

Navy-teal depths lit by the pet’s vivid lime — the snake’s own theme.

Also bundled: the claude-code, codex and opencode migration themes for easy eye-migration, plus gruvbox-dark — seven built-ins in total, selectable in either appearance slot. Pair any theme with a mascot via /pet off|red|blue|ouroboros.

Safety & proof

Trust comes from evidence

Autonomy is only useful when the result is reviewable. Gajae Code is built to leave a trail.

Plan-before-mutation

Approaches are reviewed and critiqued before any file changes, reducing wasted runs and surprise edits.

Completion evidence

Goals, revisions, checks, and verification results are recorded as durable artifacts you can read after the fact.

Isolated worktrees

Risky work runs in a dedicated git worktree that gjc creates from a new branch, so your primary checkout stays clean and the diff stays scoped.

Scoped client control

The SDK WebSocket, Coordinator MCP, and ACP expose explicit client boundaries rather than historical Bridge or RPC ingress modes.

How receipts & proof work →

Documentation

Dive deeper

Guides and references for every layer of Gajae Code.

Getting Started

Install, launch, and run your first shape → act → prove loop.

Architecture

How the CLI, agent loop, provider boundary, and tools fit together.

The Harness

Running gjc as an external runner beside your other agents.

SDK & Migration

Direct session control uses the SDK WebSocket; Coordinator MCP automates multiple sessions; ACP serves editor clients.

Skills & Agents

The four workflow skills and four role agents in detail.

Receipts & Proof

The evidence model behind durable, verifiable completion.

Give your agent a workflow it can prove

Install Gajae Code, run it beside the tools you already use, and ship work with a trail you can read.

Get Started Star on GitHub Join Discord
People

Teams

The founders building Gajae Code, oh-my-claudecode, and oh-my-codex.

Portrait of Yeachan Heo

Yeachan Heo

Creator & Founder

Creator and lead developer behind oh-my-claudecode, oh-my-codex and Gajae Code. Focused on agent orchestration, autonomous software development, and the underlying engineering architecture.

GitHub LinkedIn

Portrait of Christopher Obereder

Christopher Obereder

Co-Founder & CMO

Forbes 30 Under 30 · Growth Hacker · 250M+ users reached · 2 exits · Apple Indie Developer of the Year (2015 & 2016). Focused on growth, distribution, open-source ecosystems, partnerships and strategic development.

LinkedIn Instagram