sudoprivacy/sudocode

Rust-native CLI coding agent for hackers — terminal-native, pipe-composable, scrollback-safe. Built because Claude Code chose non-coders.

What it solves

Sudo Code provides a high-control, low-ceremony CLI coding agent for engineers who prefer working in the terminal. It aims to shorten the path from intent to verified work by offering a tool that is inspectable, composable, and model-agnostic, avoiding the "black box" feel of many AI dashboards.

How it works

Built with Rust, scode operates as an "agent unit" that can be used as a standalone CLI, a copilot, or a worker. It integrates with various model providers (Anthropic, OpenAI, xAI, Gemini) and supports multiple deployment modes via different filesystem backends (standard local FS, Nexus VFS for remote kernels, or in-process syscalls). It supports a REPL for interactive use, one-shot prompts that can be piped into other Unix tools, and a headless ACP server for integration with editors or web clients.

Who it’s for

It is designed for "hackers" and engineers who live in tmux, SSH, or IDE terminals, automate repeated tasks, and want full control over their models, credentials, and permissions.

Highlights

  • Model Agnostic: Supports a wide range of providers including Anthropic, OpenAI, xAI, and Gemini.
  • Unix-Native Design: Supports shell pipes, JSON output, and inline ANSI interaction rather than a complex TUI.
  • Composable: Can be driven by humans, orchestrators (like Hydra), or other agents via the nexus VFS plane.
  • Inspectable State: Uses file-based configuration and JSONL sessions to keep state portable and readable.
  • Extensible: Supports MCP servers, plugins, and a variety of built-in tools for bash, file editing, and web searching.

Related

  • Project
    whitelonng/mancode

    mancode is a Node‑based CLI that adds a privacy‑first, cross‑session workflow layer on top of AI coding assistants (Claude, Cursor, Copilot, etc.). It provides lightweight `solo` mode and heavier `/man`‑style planning, review, and team‑collaboration modes, persisting task references, plans, and evidence in a hidden `.mancode/` folder. The tool sanitises sensitive data before it leaves the machine, enforces engineering standards, and works with multiple agents via platform adapters. Install with `npm i -g mancode`, run `mancode init` in a project, and invoke the new `/man*` commands inside your AI assistant.

  • Project
    iamfakeguru/agent-md

    agent‑md is a lightweight framework that installs rule files, git‑hooks, and helper scripts to turn AI coding assistants (Claude Code, Codex, Cursor, Windsurf) into contract‑driven agents. It provides a single source‑of‑truth markdown file (AGENT.md) plus enforceable checks (type‑check, lint, tests, visual screenshots, state updates) that block the model until the required evidence is present.

  • Project
    smallnest/pigo

    pigo is a Go‑based CLI AI coding assistant (a clone of the pi agent) that can read/write files, run shell commands, fetch web content, and interact with many LLM providers. It works in headless script mode or an interactive REPL, supports tool‑level allow/deny lists, project trust, prompt templates, skills, plugins, self‑update, and provides a Go SDK for embedding.

  • Project
    wesammustafa/Claude-Code-Everything-You-Need-to-Know

    A practical, up‑to‑date guide for Anthropic’s Claude Code CLI, covering installation, prompt engineering, custom slash‑commands (Skills), automatic Hooks, sub‑agents, JavaScript workflows, and the Model‑Context‑Protocol for external tool integration. Includes ready‑to‑copy `.claude/` examples for teams building AI‑augmented development pipelines.

  • Project
    HKUDS/DeepCode

    An open agentic coding system that uses multi-agent systems to transform natural language and research papers into production-ready code.