Token导航 LogoToken导航TokenDH.com
研究检索只读clawhub未标认证来源可访问clear审计通过

long-project-manager长项目经理

Agent Skill

long-project-manager 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

总安装

10,639

周安装

457

GitHub Stars

公开资料未说明

下载量

3,729
OpenClaw

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

MIT-0

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:long-project-manager(长项目经理)
来源仓库:https://github.com/tianxiangma/long-project-manager
安装命令:
openclaw skills install long-project-manager
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install long-project-manager

简介

long-project-manager 创建和维护基于文件的持久工作区,适用于中断敏感的长周期任务。

  • 适合在 OpenClaw 中管理多会话或长时间运行项目时使用。
  • 通过 clawhub 安装并使用 openclaw skills install long-project-manager 命令部署。
  • 使用前需确认文件系统权限与项目状态同步机制。
  • 适用宿主包括 OpenClaw,接入前应确认版本、权限和运行环境要求。

SKILL.md

name
long-project-manager
description
Create and maintain a persistent file-based workspace for long-running, multi-session, or interruption-sensitive work. Use when the user wants to start a long-term project, continue a previous project, preserve context across days or resets, keep durable handoff notes, or avoid losing decisions and next steps in chat history. Good for engineering, research, writing, planning, and operations work.

Long Project Manager

Store long-project state in files, not only in chat.

Use the bundled templates in assets/project-template/ (Chinese) or assets/project-template-en/ (English) to create a durable project folder under projects/<project-name>/.

Project folder standard

Create or maintain these files:

  • README.md — project goal, scope, success criteria, key entry points
  • STATUS.md — current goal, current judgment, progress, blockers, next action
  • TODO.md — prioritized work queue
  • DECISIONS.md — important decisions and tradeoffs
  • LOG.md — dated work log and discoveries
  • REFERENCES.md — important files, links, commands, and outputs
  • HANDOFF.md — compact checkpoint for resuming later

Start a long project

When the user asks to create a long-running project or preserve context across sessions:

  1. Normalize a short hyphen-case project name.
  2. Create projects/<name>/.
  3. Choose a template language:

- assets/project-template/ for Chinese - assets/project-template-en/ for English

  1. Copy the full template.
  2. Fill at least:

- README.md - STATUS.md - TODO.md

  1. Tell the user the project path and the resume command or resume phrasing to use later.

Work inside the project

For every meaningful milestone, write state back to files instead of leaving it only in chat.

Update with this bias:

  • Put the latest judgment and exact next action in STATUS.md.
  • Put concrete, actionable tasks in TODO.md.
  • Put important or irreversible choices in DECISIONS.md.
  • Put dated discoveries, command results, and observations in LOG.md.
  • Put durable links, paths, and commands in REFERENCES.md.

Prefer concise, high-signal updates.

Resume a project later

Read in this order:

  1. README.md
  2. STATUS.md
  3. TODO.md
  4. DECISIONS.md
  5. LOG.md
  6. HANDOFF.md

Then summarize the project state in 3-5 bullets and continue from the first actionable next step.

Before stopping

Always leave a clean restart point:

  1. Refresh STATUS.md.
  2. Update HANDOFF.md.
  3. Make the next action explicit.
  4. Make the first file to read explicit.

Quality bar

Always preserve these four things for long tasks:

  • Current goal — what must happen now
  • Current judgment — what seems true and why
  • Open loops — what remains unresolved
  • Next action — the exact next step

If these four things are written down, the project can survive resets, context loss, and handoffs.

Good defaults

  • Keep entries short and concrete.
  • Prefer updating existing files over scattering notes across many new files.
  • Treat STATUS.md as the first-screen dashboard.
  • Treat HANDOFF.md as the restart capsule.
  • Treat chat as transient and project files as durable state.

Avoid these mistakes

  • Do not leave important project state only in chat.
  • Do not create many extra files when the standard project files are enough.
  • Do not replace durable context with vague summaries.
  • Do not stop without making the next action explicit in STATUS.md or HANDOFF.md.

适合场景

01

OpenClaw 用户查找和安装 Skill 时

02

用户想查找某类 Agent Skill 时

03

需要根据任务场景推荐可安装能力包时

04

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

补充不同宿主或平台的使用分布数据

能力 5

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

OpenClaw

73.76%
按下载量换算2,751

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills