First LLM Studio: local-first LLM studio for Apple Silicon with MLX runtimes, Compare Lab, benchmark ops, replay, and runtime telemetry.
# Add to your Claude Code skills
git clone https://github.com/ChrisChen667788/Your-First-LLM-StudioGuides for using ai agents skills like Your-First-LLM-Studio.
Last scanned: 6/18/2026
{
"issues": [
{
"type": "npm-audit",
"message": "@next/eslint-plugin-next: Vulnerability found",
"severity": "high"
},
{
"type": "npm-audit",
"message": "brace-expansion: brace-expansion: Zero-step sequence causes process hang and memory exhaustion",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "eslint-config-next: Vulnerability found",
"severity": "high"
},
{
"type": "npm-audit",
"message": "glob: glob CLI: Command injection via -c/--cmd executes matches with shell:true",
"severity": "high"
},
{
"type": "npm-audit",
"message": "js-yaml: JS-YAML: Quadratic-complexity DoS in merge key handling via repeated aliases",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "next: Next Vulnerable to Denial of Service with Server Components",
"severity": "high"
},
{
"type": "npm-audit",
"message": "picomatch: Picomatch: Method Injection in POSIX Character Classes causes incorrect Glob Matching",
"severity": "high"
},
{
"type": "npm-audit",
"message": "playwright: Playwright downloads and installs browsers without verifying the authenticity of the SSL certificate",
"severity": "high"
},
{
"type": "npm-audit",
"message": "postcss: PostCSS has XSS via Unescaped </style> in its CSS Stringify Output",
"severity": "medium"
},
{
"type": "npm-audit",
"message": "yaml: yaml is vulnerable to Stack Overflow via deeply nested YAML collections",
"severity": "medium"
}
],
"status": "WARNING",
"scannedAt": "2026-06-18T08:49:29.444Z",
"npmAuditRan": true,
"pipAuditRan": true,
"promptInjectionRan": true
}Your-First-LLM-Studio is an open-source ai agents skill for AI coding assistants such as Claude Code, Codex CLI, and ChatGPT, built by ChrisChen667788. First LLM Studio: local-first LLM studio for Apple Silicon with MLX runtimes, Compare Lab, benchmark ops, replay, and runtime telemetry. It has 101 GitHub stars.
Your-First-LLM-Studio returned warnings in SkillsLLM's automated security scan. It has no critical vulnerabilities, but review the flagged issues in the Security Report section before adding it to your workflow.
Clone the repository with "git clone https://github.com/ChrisChen667788/Your-First-LLM-Studio" and add it to your Claude Code skills directory (see the Installation section above).
Your-First-LLM-Studio is primarily written in TypeScript. It is open-source under ChrisChen667788 on GitHub, so you can review or fork the full source.
Yes. SkillsLLM lists many other AI Agents skills you can browse and compare side by side. Open the AI Agents category from the badge at the top of this page, or use the Related Skills and comparison links further down to weigh Your-First-LLM-Studio against similar tools.
No comments yet. Be the first to share your thoughts!
Requires a passing catalog security scan. Resolve the flagged issues and resubmit to enable featuring.
First LLM Studio is a local-first LLM workbench for Apple Silicon. It brings local MLX runtimes, remote API targets, Agent sessions, Compare, Fine-tune, Benchmark, model discovery, runtime recovery, release evidence, and admin monitoring into one operating surface.
It is not another chat shell. It is built for people who need to compare behavior, debug runtimes, run evals, prepare adapters, and keep local and remote model work inside one product loop.
| Route | Core workflow |
|---|---|
/agent |
Tool-enabled Agent sessions, target selection, runtime state, replay, trace review, and embedded Compare entry. |
/compare |
Route-owned Compare Studio for prompt composition, lane preview, recipe persistence, review drawer, and benchmark handoff. |
/fine-tune |
Foreground Fine-tune Studio for datasets, recipes, training, evaluation, chat adapter proof loops, export, reports, and artifacts. |
/models |
Model discovery and install verification for local/community models plus hardware-fit and risk signals. |
/benchmarks |
Benchmark run controls, progress, reports, release evidence, baselines, and regression review. |
/retrieval |
Foreground knowledge management, path import, chunk inspection, and grounded retrieval validation. |
/experiments |
Unified run/session timeline with artifact lineage, cross-feature navigation, filters, and retention controls. |
/admin |
Monitoring/configuration mirror for runtime, queues, benchmark history, provider health, guardrails, and audit timelines. |
| Version | Core capabilities |
|---|---|
v0.1 Foundation |
Established the local-first web studio, Apple Silicon/MLX gateway workflow, local + remote target catalog, runtime telemetry, and the first Agent/Admin operating split. |
v0.2 Agent + Benchmark Ops |
Added richer Agent workbench flows, Compare-style target review, replay/trace inspection, runtime recovery controls, formal benchmark operations, baselines, and regression evidence. |
v0.3 Fine-tune + Release Evidence |
Added fine-tune operation loops for evaluation, adapter chat, adapter export, and distillation starters; expanded operation history, partitioned typechecks, screenshot smoke, route smoke, and public launch assets. |
v0.4 Product IA release |
Moves /fine-tune, /compare, /models, /benchmarks, /retrieval, and /experiments into foreground product routes with feature-owned state/actions, artifact lineage and retention, dark-glass studio/workbench styling, canonical APIs, and admin narrowed toward monitoring/configuration. |
Current release: VERSION.
Local Qwen3 0.6BLocal Qwen3 4B 4-bitLocal Qwen3.5 4B 4-bitLocal Gemma 3 4B It Qat 4-bitOpenAI CodexOpenAI GPT-5.5Claude APIDeepSeek APIKimi APIGLM APIQwen API

Benchmark percentile board:

Formal milestone regression summary:

Local runtime telemetry:

22.x3.12nvm install 22
nvm use 22
npm install
cp .env.example .env.local
npm run dev
Default routes:
python3.12 -m venv .venv
source .venv/bin/activate
pip install -U pip
pip install mlx mlx-lm
python scripts/local_model_gateway_supervisor.py
Gateway health:
npm run typecheck:changed
npm run smoke:routes
npm run smoke:screenshots
Copy .env.example to .env.local and fill only the providers you want to use.
Important notes:
.env.local is ignored by git.app/ Next.js app routes and thin API transports
components/ Shared UI and compatibility shells
features/ Feature-owned routes, contracts, state, actions, and application ports
lib/agent/ Agent runtime, providers, benchmark, gateway helpers
lib/finetune/ Fine-tune store facade and split operation services
scripts/ Local gateway, runtime, verification, and release scripts
docs/ Architecture, release notes, launch notes, roadmap, and assets
modelscope/ ModelScope profile/readme metadata
public/ Public assets and social cover art
haozi667788/first-llm-studioThe ModelScope package script exports the committed Git tree so GitHub and ModelScope can stay file-identical for each synced version.
.env.local.Issues and PRs are welcome.
VERSIONdocs/releasesdocs/release-process.mdFirst LLM Studio 是一个面向 Apple Silicon 的本地优先 LLM 工作台。它把本地 MLX 运行时、远端 API 目标、Agent 会话、Compare 对比、Fine-tune 微调、Benchmark 评测、模型发现、runtime 恢复、发布证据和后台监控统一到一个产品界面里。
它不是另一个聊天壳,而是给真正需要比较模型行为、调试 runtime、跑评测、准备 adapter,并把本地/远端模型工作流收在同一个产品循环里的开发者使用。
| 路由 | 核心工作流 |
|---|---|
/agent |
带工具循环的 Agent 会话、target 选择、runtime 状态、replay、trace review,以及内嵌 Compare 入口。 |
/compare |
前台 Compare Studio,负责 prompt 编排、lane preview、recipe 持久化、review drawer 和 benchmark handoff。 |
/fine-tune |
前台 Fine-tune Studio,覆盖数据集、配方、训练、评估、adapter proof loop、导出、报告和 artifacts。 |
/models |
本地/社区模型发现、安装验证、硬件适配和风险提示。 |
/benchmarks |
Benchmark run controls、进度、报告、发布证据、baseline 和回归审阅。 |
/retrieval |
前台知识管理、路径导入、chunk 检查和 grounded retrieval 验证。 |
/experiments |
统一 Session/Run 时间线、artifact lineage、跨功能导航、筛选和保留策略。 |
/admin |
Runtime、队列、benchmark 历史、provider health、guardrails 和 audit timeline 的监控/配置镜像。 |
| 版本 | 核心功能 |
|---|---|
v0.1 基础版 |
建立本地优先 Web Studio、Apple Silicon/MLX 网关工作流、本地 + 远端 target catalog、runtime telemetry,以及 Agent/Admin 的第一版操作分层。 |
v0.2 Agent + Benchmark 运维 |
增强 Agent 工作台、Compare 式 target review、replay/trace 检查、runti |