Builderoom
Get resource

A metacognitive framework and skill plugin providing Rust development assistance for AI coding assistants

MITClaude CodeText InputCode InputCode Output
安装 / 调用方式bash
# Skills-only 模式 (NPX)
npx skills add actionbook/rust-skills

# Skills-only 模式 (CoWork CLI)
cargo install cowork
cowork install actionbook/rust-skills

# Plugin 模式 (Marketplace)
/plugin marketplace add actionbook/rust-skills
/plugin install rust-skills@rust-skills

# Plugin 模式 (Local)
git clone https://github.com/actionbook/rust-skills.git
claude --plugin-dir /path/to/rust-skills

Product facts

Type
AI Skill & Agent
Category
AI & Machine Learning › Prompt Engineering & Skill Management
Pricing
Open source
Platforms
Claude Code
License
MIT
Launched
Sep 26, 2026
Listed
Sep 26, 2026
Website
github.com

Overview

GitHubGitHub

**Rust Skills** is a Rust development assistance system for AI coding assistants, currently at version 2.0.9. It primarily serves as a plugin for Claude Code, while also supporting integration with other skills-compatible coding agents (such as Vercel AI) in Skills-only mode. ### What Problem It Solves Traditional AI assistants handling Rust code often offer superficial fixes (e.g., suggesting `.clone()` for an E0382 error) while ignoring domain constraints and architectural soundness. Rust Skills introduces a metacognitive framework that guides the AI through multiple layers of cognitive reasoning before answering, producing architecture-level solutions aligned with domain correctness. ### Core Mechanism: Metacognitive Framework The system uses a three-layer cognitive model, requiring the AI to perform hierarchical tracing rather than answering directly: - **Layer 3: Domain Constraints (WHY)**: Analyzes how domain rules dictate design choices, e.g., financial systems require immutable, auditable data. - **Layer 2: Design Choices (WHAT)**: Determines design patterns and architectural decisions, e.g., using `Arc<T>` to share immutable data. - **Layer 1: Language Mechanics (HOW)**: Maps to Rust language features and compiler rules, e.g., treating E0382 as a symptom of an ownership design issue. Built-in routing rules automatically decide the entry layer and tracing direction based on user input signals (e.g., E0xxx errors, performance issues, specific domain requirements). ### Key Capabilities - **Dynamic Skill Generation**: Automatically generates corresponding assistance skills based on project `Cargo.toml` dependencies. - **Real-time Information Retrieval**: Fetches the latest Rust versions and crate info via background agents (relying on actionbook or agent-browser). - **Domain Extensions**: Built-in support for FinTech, ML, Cloud-Native, IoT, Embedded, Web, CLI, and more. - **Coding Standards**: Provides comprehensive queries for Rust coding conventions and best practices. - **Skill System**: Includes 31 skills covering main routing (`rust-router`), information retrieval (`rust-learner`), coding standards (`coding-guidelines`), plus specialized skills for ownership, resource management, mutability, zero-cost abstractions, type-driven design, error handling, concurrency (m01-m07), and design-layer skills (m09-m15). ### Installation Two installation modes are supported: 1. **Plugin Mode**: For Claude Code, includes full features like hooks, agents, and automatic metacognitive triggers. Install via Marketplace commands or local cloning. 2. **Skills-only Mode**: For any skills-compatible coding agent. Install via NPX, CoWork CLI, or manual file copying. This mode excludes hooks; you must manually invoke `/rust-router` or specific skills, and background agents automatically fall back to inline execution. ### Who It's For Engineers using AI coding assistants for Rust development, especially those who need to ensure architectural soundness and correct ownership design in complex business domains such as finance, embedded systems, and cloud-native applications.

Key features

  • Metacognitive Framework: Uses a three-layer cognitive model (Domain → Design → Mechanics) to guide the AI toward deep architectural reasoning instead of superficial fixes
  • Real-time Information Retrieval: Fetches the latest Rust versions and crate info via background agents
  • Dynamic Skill Generation: Automatically generates corresponding assistance skills based on project Cargo.toml dependencies
  • Multi-Domain Extension Support: Built-in extensions for FinTech, ML, Cloud-Native, IoT, Embedded, Web, CLI, and more
  • Coding Standards Reference: Provides comprehensive Rust coding conventions and best practices
  • Smart Routing: Automatically matches the entry layer and invokes the corresponding skill based on user-provided error codes or problem types
  • Ownership & Lifetime Analysis: Analyzes data ownership and borrowing patterns for errors like E0382 and E0597
  • Concurrency Model Guidance: Distinguishes CPU-bound from I/O-bound scenarios, guiding the use of async, Send, Sync, and threads
  • Dual-Mode Installation: Supports full-featured Plugin Mode for Claude Code and Skills-only Mode compatible with other agents
  • Inline Fallback Logic: When agent files are unavailable, skills can execute directly using built-in tools

Releases

Notify me

Initial · Rust Skills

Q&A

What is Rust Skills?AI-generated
Rust Skills is an AI-powered Rust development assistance system, primarily used as a Claude Code plugin. Through its metacognitive framework, it guides the AI to reason across three cognitive layers—domain, design, and language mechanics—before answering Rust questions, delivering architecture-level solutions that respect domain constraints rather than just superficial fixes.
How does the metacognitive framework work?AI-generated
The framework consists of three layers: Layer 3 focuses on domain constraints (WHY), Layer 2 on design choices (WHAT), and Layer 1 on language mechanics (HOW). The system decides the entry layer based on user signals. For example, an E0xxx error triggers upward tracing from Layer 1, while domain-specific application development triggers downward tracing from Layer 3.
What installation methods are supported?AI-generated
Two modes are supported. Plugin Mode is for Claude Code, installable via Marketplace or local cloning, offering full features including hooks and automatic metacognitive triggers. Skills-only Mode works with any skills-compatible coding agent, installable via NPX, CoWork CLI, or manual copying, but requires manual skill invocation.
Can it be used on platforms other than Claude Code?AI-generated
Yes. In Skills-only mode, it works with any skills-compatible coding agent, including Vercel AI. Additionally, the repository provides installation documentation for OpenCode and Codex platforms.
How many skills does it include?AI-generated
According to the feature comparison table, Rust Skills includes 31 skills in total. These cover core routing, information retrieval, coding standards, and specialized skills for ownership, resource management, mutability, zero-cost abstractions, type-driven design, error handling, and concurrency.
What permission configuration is required to run background agents?AI-generated
Background agents require permission to run agent-browser. Create a .claude/settings.local.json file in your project and add "Bash(agent-browser *)" to the permissions.allow array.
What is the difference between Skills-only Mode and Plugin Mode?AI-generated
Both include all 31 skills and background agent capabilities (with inline fallback in Skills-only mode). However, Plugin Mode supports automatic metacognitive triggers and hook-based routing, and is easier to update. Skills-only mode lacks automatic triggers and hook routing, but works with other coding agents.
More
Get resource
Bookmark

This product has no owner yet

It was listed by the platform or its author left. If it is your work, prove your relationship to become the owner.

Claim this product

Alternatives

The product content on this page was translated by AI from the author's original.