JiegeOSBench — Development skill for Claude Code
Benchmarking a single‑prompt Rust OS kernel implementation, comparing Claude Code Fable, Opus, and Sonnet.
How to install JiegeOSBench
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open wangrunji0408/JiegeOSBench and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What JiegeOSBench does
Benchmarking a single‑prompt Rust OS kernel implementation, comparing Claude Code Fable, Opus, and Sonnet.
Alternatives in Development
- OpenAI Codex CLI — (Rust implementation) 67.8k ★
- Quick Plan — Create a detailed implementation plan based on the user's requirements provided through the USER_PROMPT variab 3.4k ★
- Claudes C Compiler — Claude Opus 4.6 wrote a dependency-free C compiler in Rust, with backends targeting x86 (64- and 32-bit), ARM 2.6k ★
README
JiegeOSBench
[中文版](./README-CN.md)
A benchmark evaluating how well LLM coding agents can autonomously implement a RISC‑V OS kernel from scratch — running an unmodified Linux nginx binary on QEMU, serving HTTP from the host.
Prompt: You are the AI-Jiege. Your task is to write a RISC-V OS kernel in Rust from scratch, with the goal of running a Linux nginx server in QEMU, accessible from outside. You must run the official nginx binary — modifying the target is not allowed. Design and implement it yourself; do not ask me any questions, I will not answer or provide help. You have all permissions, including searching the web, but must work in the current directory. Keep working until the goal is achieved.
| # | Model | Effort | Harness | Duration | Context | Cost | Branch | Tier |
|---|---|---|---|---|---|---|---|---|
| 🏅 | GPT-6 Astra | High | Codex | 6min | 83K | $4 | gpt-6-astra | Jiege |
| 🥈 | Claude Fable 5 | High | CC | 38min | 155K | $21 | fable-5 | Super Jiege |
| 🥉 | GPT 5.6 Sol | High | Codex | 36min | 222K | $14 | gpt-5.6-sol | Super Jiege |
| 4 | Claude Fable 5.1 | High | CC | 58min | 516K | $34 | fable-5.1 | Smart Jiege |
| 5 | Claude Opus 5 | High | CC | 67min | 334K | $26 | opus-5 | Smart Jiege |
| 6 | Claude Opus 4.7 | — | CC | 65min | — | — | opus-4.7 | Smart Jiege |
| 7 | DeepSeek V4 Pro | High | DSH | 108min | 503K | $0.86 | deepseek-v4-pro | Smart Jiege |
| 8 | Kimi K3 | High | CC | 2h 19min | 270K | $11 | [kimi-k3](https://github.com/wangrunji0408 |
Related Skills
Claude Fable 5 Prompt Strategies
Use when composing prompts targeting Claude Opus 5 (Claude 디폴트), Fable 5, Opus 4.8, or Sonnet 5 — 최신 Claude 프롬
Advisor Tool Demo
Demo app for Anthropic's Advisor Tool (beta) — customer support agent comparing cost/quality tradeoffs across
Claude Code Game Builder
Claude Code skill: multi-agent build → capture → blind critic → fix loops that push a game toward a reference-
Caveman Micro
6-line caveman micro prompt (85 tokens) that outperformed the original 552-token skill. Benchmark on Claude So
Triton GPU Kernel Optimization
Write high-performance tiled Triton GPU kernels with autotune, grouped tile ordering, stride-based addressing,
Fable OS
An agentic operating system where the kernel is controlled directly by Claude
Related Agents
Bench Implementer
Implements a single fable-bench packet from a brief. Sonnet 5, builds with the six fable skills loaded. Use fo
Fable Architect
Plan-only Fable 5.1 pass from an Opus or Sonnet main. Use for architecture, ambiguous root cause, or a design
Escalation Reviewer Opus
Second-level escalation code reviewer (Opus) used ONLY under ORCHESTRATOR section 6, when the Sonnet escalatio
