AI Review

Product · Open Source

yottacode

Sovereign AI coding from your terminal.

An open-source coding agent for developers and teams that need control over source code, tools, models, permissions, memory, and execution environment.

License MITPlatforms Linux · macOSRuntime Single Go binary

Feature map

Everything you need to let an agent work safely in a real repo.

Memory that learns you

Plain-text, grep-able memory: you own USER.md, the agent curates the rest, and project writes are approval-gated. Per-turn retrieval surfaces only what is relevant, so context gets sharper the more you use it.

Use any model, switch any time

Native adapters for Anthropic, OpenAI, Gemini, xAI, local Ollama, ChatGPT, GitHub Copilot, and OpenAI-compatible gateways. Swap mid-session without lock-in or hidden defaults.

A built-in security policy

First-launch trust prompts, syntax-highlighted approval diffs, allow / ask / deny project rules, and write-path validation that firewalls secret-bearing paths like .env and ~/.ssh.

Deep GitHub integration

A typed GitHub adapter with no gh CLI required. Read, open, update, and review PRs, triage issues, and take an issue through research, plan, branch, implementation, tests, and draft PR.

Plan first, then let it run

Plan mode investigates read-only and drafts a plan you approve. Approving drops into auto mode while bash and commits stay gated, and the agent can never escalate its own permissions.

A deep, repo-aware tool surface

50 built-in tools for reads, writes, search, git, bash, tests, working-plan tracking, and repository workflows — each with an explicit approval policy.

Rewind and resume

Every prompt auto-checkpoints the conversation and pre-edit files. A double-tap of Esc rolls back either or both, sessions save atomically each turn, and /recall searches across them.

Delegate to typed subagents

Hand research, code search, planning, and verification to subagents in their own context windows. The parent sees only the final answer, and custom agents live under .yottacode/agents/.

Reusable skills, on demand

Built-in playbooks for TDD, security audit, performance profiling, remote ops, and more load only when relevant. Install your own from a path, URL, or GitHub shorthand.

Work in parallel

Run sessions in isolated git worktrees so two agents can edit the same repo without colliding. A per-repo .worktreeinclude carries gitignored configs into each one.

Make it your own

Drop markdown files into .yottacode/commands/ for custom slash commands with argument substitution and file references, then connect MCP servers for tools beyond the built-ins.

Built for the terminal

Inline rendering keeps scrollback intact with markdown output, a Tab-completing slash palette, and image paste. Script the same core in one-shot mode from a static Go binary.

Get Started

Try yottacode on a real repository.

Install yottacode and evaluate it against real engineering work. Current install instructions, documentation, releases, and checksums live on the product site and GitHub repository.