Skip to content
Back to search

code-assistant

An LLM-powered, autonomous coding assistant. Also offers an MCP and ACP mode.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/stippi/code-assistant with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/stippi/code-assistant'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/stippi--code-assistant/badge.svg)](/listings/stippi--code-assistant)

HTML

<a href="/listings/stippi--code-assistant"><img src="/api/listings/stippi--code-assistant/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

These steps run a script from the network that findsafeskills does not scan.

The quickest way to try it is a prebuilt download — no toolchain required:

1. Grab the latest build from the [**Releases page**](https://github.com/stippi/code-assistant/releases/latest): - **macOS:** `Code-Assistant-macos-aarch64.app.zip` (Apple Silicon) or `Code-Assistant-macos-x86_64.app.zip` (Intel) - **Linux:** `code-assistant-linux-x86_64.zip` - **Windows:** `code-assistant-windows-x86_64.zip` 2. Launch it. On first start, code-assistant opens its **Settings** screen. Pick a provider (there are one-click suggestions for the common ones), paste an API key, and choose a model — that's it. 3. Add your project folder from within the app and start a chat.

No hand-editing of JSON files required to get going. (You still can, if you prefer — see the [Configuration guide](docs/configuration.md).)

<details> <summary>Build from source instead</summary>

# Install the Rust toolchain (macOS/Linux)
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh

# On Linux (Debian/Ubuntu), install the system libraries gpui needs:
sudo apt-get install -y --no-install-recommends \
    pkg-config build-essential libssl-dev libzstd-dev \
    libfontconfig-dev libwayland-dev libx11-xcb-dev \
    libxkbcommon-x11-dev libasound2-dev libvulkan1

# On macOS, install the metal toolchain:
xcodebuild -downloadComponent MetalToolchain

# Clone and build
git clone https://github.com/stippi/code-assistant
cd code-assistant
cargo build --release

The binary lands at `target/release/code-assistant`. See the [Configuration guide](docs/configuration.md) for building a self-contained macOS `.app` bundle.

</details>

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 180 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # Code Assistant
  2. [![CI](https://github.com/stippi/code-assistant/actions/workflows/build.yml/badge.svg)](https://github.com/stippi/code-assistant/actions/workflows/build.yml)
  3. [![Trust Score](https://archestra.ai/mcp-catalog/api/badge/quality/stippi/code-assistant)](https://archestra.ai/mcp-catalog/stippi__code-assistant)
  4. [![MCP Toplist](https://mcptoplist.com/badge/mcp.so%2Fcode-assistant%2Fstippi.svg)](https://mcptoplist.com/server/mcp.so%2Fcode-assistant%2Fstippi)
  5. An open-source AI coding agent, written in Rust, with a native GUI, a terminal
  6. mode, and integrations for editors and MCP clients. It runs an autonomous agent
  7. loop over your codebase — reading, searching, editing files, and running
  8. commands — while keeping you in the loop about what it is actually doing.
  9. ## Getting started
  10. The quickest way to try it is a prebuilt download — no toolchain required:
  11. 1. Grab the latest build from the [**Releases page**](https://github.com/stippi/code-assistant/releases/latest):
  12. - **macOS:** `Code-Assistant-macos-aarch64.app.zip` (Apple Silicon) or `Code-Assistant-macos-x86_64.app.zip` (Intel)
  13. - **Linux:** `code-assistant-linux-x86_64.zip`
  14. - **Windows:** `code-assistant-windows-x86_64.zip`
  15. 2. Launch it. On first start, code-assistant opens its **Settings** screen.
  16. Pick a provider (there are one-click suggestions for the common ones), paste
  17. an API key, and choose a model — that's it.
  18. 3. Add your project folder from within the app and start a chat.
  19. No hand-editing of JSON files required to get going. (You still can, if you
  20. prefer — see the [Configuration guide](docs/configuration.md).)
  21. <details>
  22. <summary>Build from source instead</summary>
  23. ```bash
  24. # Install the Rust toolchain (macOS/Linux)
  25. curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
  26. # On Linux (Debian/Ubuntu), install the system libraries gpui needs:
  27. sudo apt-get install -y --no-install-recommends \
  28. pkg-config build-essential libssl-dev libzstd-dev \
  29. libfontconfig-dev libwayland-dev libx11-xcb-dev \
  30. libxkbcommon-x11-dev libasound2-dev libvulkan1
  31. # On macOS, install the metal toolchain:
  32. xcodebuild -downloadComponent MetalToolchain
  33. # Clone and build
  34. git clone https://github.com/stippi/code-assistant
  35. cd code-assistant
  36. cargo build --release
  37. ```
  38. The binary lands at `target/release/code-assistant`. See the
  39. [Configuration guide](docs/configuration.md) for building a self-contained
  40. macOS `.app` bundle.
  41. </details>
  42. ## What makes it different
  43. A handful of things we care about:
  44. - **A UI you can actually follow.** Instead of a terse "Called 5 tools",
  45. code-assistant shows which tools ran and what each of them
  46. handed back to the model as context. When it goes off the rails, you can see
  47. why.
  48. - **Format-on-save that stays token-efficient.** When your project auto-formats
  49. code, an agent's mental picture of a file goes stale and its follow-up edits
  50. start failing for non-obvious reasons. code-assistant runs your formatter and
  51. then reconciles the model's own edits with the formatted result — without
  52. wasting tokens on re-reading files. See
  53. [docs/format-on-save-feature.md](docs/format-on-save-feature.md).
  54. - **Transparent file encoding & line endings.** It reads a file however it is
  55. stored (encoding, BOM, CRLF/LF), gives the model clean text, and writes it
  56. back the way it was — so it never silently rewrites your line endings.
  57. - **Searches and reads documents, not just code.** Word, Excel, PowerPoint,
  58. PDF and more are consumed automatically as Markdown, so you can point the
  59. agent at real-world documents alongside your source.
  60. ## Features
  61. - **Multiple LLM providers:** Anthropic, OpenAI, Google Vertex AI, Ollama,
  62. OpenRouter, SAP AI Core, Groq, Cerebras, Mistral, and more.
  63. - **Four ways to work:** a native GUI (built on Zed's GPUI), a terminal mode, a
  64. headless MCP server, and an ACP agent for editors like [Zed](https://zed.dev).
  65. - **Adaptive tool syntax:** native function calling, XML tags, or triple-caret
  66. blocks — chosen per session to fit the model.
  67. - **Real-time streaming** with smart filtering that blocks unsafe tool
  68. combinations (e.g. editing a file before reading it).
  69. - **Sessions per project** with branching, persistent state, and draft messages
  70. with attachments.
  71. - **Sub-agents, permission tiers, a command sandbox,** and automatic context
  72. compaction when the window fills up.
  73. - **MCP client mode:** plug in external Model Context Protocol servers and use
  74. their tools.
  75. - **Browser sessions:** the agent drives a real browser (navigate, read, click,
  76. type) to test web apps, with a human-in-the-loop login handoff so it acts as
  77. you without ever seeing your credentials.
  78. - **Skills:** reusable, task-specific playbooks the agent can load on demand.
  79. - **Auto-loaded guidance:** picks up `AGENTS.md` (or `CLAUDE.md`) from your
  80. project root to align with repo-specific instructions.
  81. ## Interfaces
  82. ```bash
  83. code-assistant # native GUI (default)
  84. code-assistant --tui # terminal interface
  85. code-assistant acp # ACP agent for editors like Zed
  86. code-assistant server # headless MCP server (e.g. Claude Desktop)
  87. ```
  88. Any mode can take an initial task: `code-assistant --task "Explain this codebase"`.
  89. <details>
  90. <summary>Connect to Zed (ACP)</summary>
  91. Add to your Zed settings:
  92. ```json
  93. {
  94. "agent_servers": {
  95. "Code-Assistant": {
  96. "command": "/path/to/code-assistant",
  97. "args": ["acp", "--model", "Claude Sonnet 4.5"],
  98. "env": { "ANTHROPIC_API_KEY": "sk-ant-..." }
  99. }
  100. }
  101. }
  102. ```
  103. See [Zed's docs on custom agents](https://zed.dev/docs/ai/external-agents#add-custom-agents).
  104. </details>
  105. <details>
  106. <summary>Connect to Claude Desktop (MCP)</summary>
  107. In Claude Desktop settings (**Developer** tab → **Edit Config**):
  108. ```json
  109. {
  110. "mcpServers": {
  111. "code-assistant": {
  112. "command": "/path/to/code-assistant",
  113. "args": ["server"],
  114. "env": { "SHELL": "/bin/zsh" }
  115. }
  116. }
  117. }
  118. ```
  119. </details>
  120. ## Configuration
  121. For most people the in-app **Settings** screen is enough — it manages providers,
  122. models, MCP servers and skills for you. Everything can also be configured via
  123. JSON files, and there are more advanced options (project setup, sandbox modes,
  124. tool syntax, session recording, CLI flags):
  125. **→ See the [Configuration guide](docs/configuration.md).**
  126. ## Contributing
  127. Contributions are welcome! The codebase is a decent tour of async Rust, AI agent
  128. architecture, and cross-platform UI. If something about the agent's behaviour
  129. annoys you, that is exactly the kind of detail this project cares about: please
  130. open an issue.
  131. ## Roadmap
  132. Not really a roadmap — just a few directions, in no particular order:
  133. - **Block replacing in stale files:** detect early when the model is about to
  134. `replace_in_file` a file that changed since it last read it, and reject with a
  135. helpful message.
  136. - **Compact tool-use failures:** strip failed tool calls / mismatched search
  137. blocks from the history to save tokens.
  138. - **Memory tools:** help the agent build up a knowledge base for a project.
  139. - **Tighter sandboxing:** restrict tool access to git-tracked files within the
  140. project.
  141. - **Fuzzy matching of search blocks:** reduce the re-reads caused by failed
  142. exact matches.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.