npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@ciqtek/xigua

v0.1.12

Published

CLI-first local knowledge workbench agent with a reusable runtime core.

Readme

Xigua Agent

Xigua Agent 是一个受 claude-code 启发、可在本机运行、以 CLI 为优先入口的知识工作台智能体。当前实现采用可复用的 runtime core 与全屏终端 shell 分层设计。

当前状态

V1 当前聚焦于:

  • 全屏 TUI 交互
  • OpenAI / OpenAI-compatible 后端
  • 本地文件、搜索、shell、网页工具
  • 正式的多智能体任务运行时
  • 会话持久化与恢复
  • 权限控制
  • 权限模式与行为模式
  • 本地 skills、agents 与 MCP 集成

pslogsattachkilldaemonremote-controldesktopmobile 等非核心公开入口在当前版本中以兼容占位形式存在。

快速开始

从 npm 安装:

npm install -g @ciqtek/xigua
xigua doctor
OPENAI_API_KEY=your_key xigua

本地源码运行:

pnpm install
pnpm doctor
OPENAI_API_KEY=your_key pnpm dev

或者先构建 CLI:

pnpm build
node dist/cli/index.js doctor
OPENAI_API_KEY=your_key node dist/cli/index.js

配置

首次运行时,Xigua Agent 会创建:

  • 配置文件:~/.xigua-agent/config.json
  • 会话目录:~/.xigua-agent/sessions

默认 provider 配置使用 OPENAI_API_KEY,默认模型为 gpt-4.1-mini

你可以在 TUI 内查看和更新配置:

  • /config show
  • /config path
  • /config set model gpt-4.1
  • /config set baseURL http://localhost:8000/v1
  • /config set apiKeyEnv MY_OPENAI_KEY
  • /config set permissionMode bypass
  • /config set behaviorMode chat

当前模式模型分为两条轴:

  • 权限模式:ask | auto | bypass
  • 行为模式:chat | plan

其中:

  • bypass 适合受控远程运行环境
  • plan 会切换为规划工作流,默认只暴露只读/分析类工具,并通过 Plan 工具把计划沉淀到 workspace/.xigua/plans/

Slash Commands

  • /help
  • /clear
  • /model
  • /config
  • /tools
  • /permissions
  • /mode
  • /status
  • /resume
  • /doctor
  • /mcp
  • /agents
  • /skills
  • /plan
  • /tasks

其中与模式和规划直接相关的命令包括:

  • /permissions mode <ask|auto|bypass>
  • /mode <chat|plan>
  • /plan
  • /plan show
  • /plan approve
  • /plan revise
  • /plan cancel

文档

完整文档索引见 docs/README.md

关键入口:

  • docs/requirements/agent/agent-001-cli-office-agent-v1-requirement.md
  • docs/plans/agent/agent-001-cli-office-agent-v1-design-v1.md
  • docs/plans/agent/agent-001-cli-office-agent-v1-plan-v1.md
  • docs/tests/agent/agent-001-cli-office-agent-v1-test-report-20260401.md
  • docs/summary/agent-001-cli-office-agent-v1-summary-20260401.md
  • docs/requirements/agent/agent-002-multi-provider-model-requirement.md
  • docs/plans/agent/agent-002-multi-provider-model-plan-v1.md
  • docs/requirements/agent/agent-004-phase-0-task-runtime-requirement.md
  • docs/plans/agent/agent-004-phase-0-task-runtime-design-v1.md
  • docs/plans/agent/agent-004-phase-0-task-runtime-plan-v1.md
  • docs/tests/agent/agent-004-phase-0-task-runtime-test-report-20260402.md
  • docs/summary/agent-004-phase-0-task-runtime-summary-20260402.md
  • docs/requirements/agent/agent-006-permission-and-behavior-modes-requirement.md
  • docs/plans/agent/agent-006-permission-and-behavior-modes-design-v1.md
  • docs/plans/agent/agent-006-permission-and-behavior-modes-plan-v1.md
  • docs/tests/agent/agent-006-permission-and-behavior-modes-test-report-20260402.md
  • docs/summary/agent-006-permission-and-behavior-modes-summary-20260402.md
  • docs/pmm/pmm-002-xigua-agent-vs-claude-code-gap-and-readiness-20260404.md
  • docs/pmm/pmm-004-xigua-agent-claude-openclaw-comprehensive-comparison-20260405.md

任务运行时

Agent 工具现在会创建正式 task,而不是只返回一段字符串摘要。

当前支持:

  • 父 session 记录 task 元数据
  • 子任务使用独立 child session 保存 transcript
  • /status 显示任务统计
  • /tasks list
  • /tasks show <task-id>
  • /tasks open <task-id>

这意味着主 agent、subagent 和更深一层的 child agent 都可以通过同一套 session/task 模型管理,而不是把多智能体行为隐藏在一次普通工具调用后面。

权限与规划模式

当前 xigua-agent 已支持:

  • 正式 bypass 权限模式
  • chat / plan 行为模式
  • plan 文档存储与审批流
  • Plan / SubmitPlan 双阶段规划工具流

关键行为:

  • plan 模式下默认只暴露 Read / Glob / Grep / WebFetch / WebSearch / Skill / Agent / Plan
  • 规划完成后,模型应通过 SubmitPlan 把状态推进到审批阶段
  • Plan 工具会把计划落盘到 workspace/.xigua/plans/<session-id>.md
  • 计划进入 ready_for_approval 后,TUI 会显示审批面板
  • 批准后会自动退出 plan 并回到 chat