@aperant/framework

中文文档
0.35.0发布于 2026/11/01AGPL-3.0-onlyTypeScriptES Module近 30 天下载 1万

AI coding framework — composable skills for planning, executing, verifying, and reviewing code with any LLM provider. Works as Claude Code commands, Codex tasks, or programmatically.

原始使用文档
文档目录5 节
  1. Install
  2. Usage
  3. Available commands
  4. Auto-spar plans
  5. License

Aperant Framework

npm License: AGPL v3

Work in progress. This is the foundation that will run the Aperant ecosystem — an agentic IDE covering the full software lifecycle. Join the waitlist at aperant.com.

Install

npx @aperant/framework init .

For non-interactive install or re-init across common runtimes:

npx @aperant/framework init . --claude --codex --pi --yes

Usage

Just type /apt and describe what you want in natural language. The router classifies your input and dispatches to the right skill.

/apt add OAuth login with Google
/apt the chat scroll is jittery — debug it
/apt help me think through the new pricing model
/apt typo in the footer — wrong year
/apt review my PR

You can also call any skill directly (e.g. /apt:plan ..., /apt:debug ...).

Runtime command surfaces differ slightly:

  • Claude: /apt:plan
  • Codex: /apt-plan
  • Pi: /apt, /apt:plan via .pi/prompts/

Available commands

StageCommandPurpose
Route/aptSmart router — natural-language dispatch to any skill
Discovery/apt:scanBuild/refresh .aperant/features/ registry
Discovery/apt:zoom-outOne-page architectural orientation for an unfamiliar area
Discovery/apt:improve3-lens scan for refactor opportunities (DEEPENING / INTERFACE-DESIGN / LANGUAGE)
Discovery/apt:debugSystematic debugging with persistent state across context resets
Planning/apt:discussBrainstorm a vision or lock ambiguous decisions before planning
Planning/apt:designPre-coding wireframe → critique → handoff to /apt:plan
Planning/apt:mockupVisual HTML+Tailwind mockup loop with parallel variants
Planning/apt:planInvestigate codebase, produce spec.md + implementation_plan.json
Planning/apt:roundtable10-agent structured debate with supermajority voting
Planning/apt:sparBidirectional adversarial sparring with a second CLI provider
Planning/apt:stress-testScenario simulation to catch operational gaps before execution
Planning/apt:prototypeQuestion-driven throwaway prototyping (LOGIC + UI branches)
Build/apt:quickFast single-shot task — skip planning, go straight to implementation
Build/apt:executeBuild code from a plan with atomic commits and progress tracking
Build/apt:runFull pipeline: plan → execute → verify → review in one command
Verify/apt:verify4-dimension QA scoring (correctness / completeness / quality / coverage)
Verify/apt:verify-proofVisual proof verification — screenshots, video, HTML report
Verify/apt:review6-pass inline code review with fix-all pipeline
Verify/apt:pr-reviewMulti-agent PR review — 6-8 parallel specialists (i18n + cross-platform engage conditionally) + validator + fixer + self-reviewer
Ship/apt:shipCreate PR with QA scores, decision traceability, commit-to-subtask mapping
Ship/apt:watch-ciPost-ship CI watcher — polls PR checks, auto-fixes red, pushes, reschedules
Ship/apt:close-taskPost-merge closer — confirm merged, flip milestone, GC state
Session/apt:pauseHuman-initiated handoff — save CONTINUE artifacts when you need to stop
Session/apt:resumeRestore context from paused session and route to next action
Docs/apt:docsGenerate/update documentation verified against the actual codebase
Docs/apt:create-docsPersona-first product docs from PROJECT.md + personas.json + features registry
Docs/apt:diagramRender a DiagramScene to .excalidraw file or live Excalidraw MCP canvas
Project/apt:bootstrapGenerate/refresh PROJECT.md (product-context source of truth)
Project/apt:personasDiscover 3–5 product personas grounded in code + PROJECT.md
Project/apt:roadmapManage per-scope cards, milestones, owners, ROADMAP.md rendering
Project/apt:triageMove a task through the 5-state machine
Framework/apt:setupChange preferences (verification style, multi-model, parallelization, branching)
Framework/apt:updateRefresh installed skills / agents / hooks across every runtime

Auto-spar plans

Have /apt:run challenge every plan with a second LLM before you approve. In .aperant/config.local.json:

{ "spar": { "enabled": true, "partner": "codex" } }

partner = codex | claude | gemini (not your current CLI). Fires after the plan stage (STANDARD) or plan + execute (DEEP); QUICK is exempt. Or run /apt:setup → Pipeline → Sparring Auto-Fire.

License

AGPL-3.0-only. See LICENSE.

版本与发布记录

@aperant/framework 共发布 46 个版本,首次发布于 2026/05/13;过去 12 个月发布了 46 个版本。

版本发布日期距今
0.35.0latest2026/11/016 天前
0.34.12026/10/2612 天前
0.34.02026/10/2216 天前
0.33.02026/10/1523 天前
0.30.02026/09/261 个月前
0.29.32026/09/042 个月前

其他发布通道:next → 0.13.0-next.0。

关于 @aperant/framework 的常见问题

@aperant/framework 是什么?

AI coding framework — composable skills for planning, executing, verifying, and reviewing code with any LLM provider. Works as Claude Code commands, Codex tasks, or programmatically.

如何安装 @aperant/framework?

在项目目录运行 npm install @aperant/framework,或使用 pnpm add / yarn add / bun add。它还提供命令行工具 apt-tsc、apt-tools、framework、apt-proof-video,可以用 npx apt-tsc 直接运行。

@aperant/framework 支持 TypeScript 吗?

支持。包内自带 TypeScript 类型声明(package.json 声明了 types 入口),安装后即可获得类型提示,无需再安装 @types 包。

@aperant/framework 支持 ESM 还是 CommonJS?

以 ES Module 形式发布,推荐使用 import 引入。在 CommonJS 项目中,Node.js 20.19+ / 22.12+ 可以直接 require() 不含顶层 await 的 ES Module,更早的版本需要改用动态 import()。

@aperant/framework 需要什么版本的 Node.js?

package.json 的 engines 字段要求 Node.js >=18.0.0。版本不符时 npm 默认只给出警告,开启 engine-strict 后会拒绝安装。

@aperant/framework 可以商用吗?

@aperant/framework 使用 AGPL-3.0-only 许可。最严格的 Copyleft:除分发外,通过网络向用户提供基于它的服务也需要开源对应源码。闭源 SaaS 使用前务必评估。以上为概括,不构成法律意见,请以许可原文为准。

@aperant/framework 的下载量有多少?

根据 npm 官方下载统计,@aperant/framework 在 2026/10/06 – 2026/11/04 这 30 天内被下载 10,273 次,日均约 342 次。下载量会把 CI 构建、镜像同步和作为其他包依赖的安装一并计入,适合横向比较热度,不能直接等同于使用人数。

@aperant/framework 还在维护吗?

最新版本 0.35.0 发布于 2026/11/01(6 天前);至今共发布 46 个版本,其中过去 12 个月发布了 46 个;首次发布于 2026/05/13。发布频率只是参考信号:功能稳定的工具可能很少需要更新,建议结合源码仓库的 issue 和提交记录判断。