Token导航 LogoToken导航

Agent Skill · 研究检索

map-project(地图项目)

map-project 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 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。

map-project 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为需要联网;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 39,901。

总安装 1,350周安装 58GitHub Stars 公开资料未说明下载 473核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

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:map-project(地图项目)
来源仓库:https://github.com/ulpi-io/skills
仓库路径:skills/map-project
安装命令:
npx skills add https://github.com/ulpi-io/skills --skill map-project
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 npx skills 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。

skills.shnpx skills
npx skills add https://github.com/ulpi-io/skills --skill map-project

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

map-project 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 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。它支持通过关键词、任务场景或来源线索进行信息检索与筛选。可通过 npx skills add 命令从指定 GitHub 仓库安装并使用。

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。适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

来源仓库内路径:skills/map-project

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/ulpi-io/skills --skill map-project;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它需要需要联网,先在测试目录或测试账号里跑,确认没问题再放到正式项目。

常见问题

6 条

map-project 怎么安装?

在终端执行 npx skills add https://github.com/ulpi-io/skills --skill map-project。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

map-project 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

map-project 需要什么权限?

权限级别是需要联网。该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

map-project 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

map-project 用的人多吗?

公开数据是累计安装 1,350 次,近一周安装 58 次,下载 473 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

map-project 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
map-project

地图项目 · 本页

1,350—需要联网未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Map Project

Inputs

  • $request: Optional target scope or refresh note such as api only, after auth refactor, or update the docs for the Next.js app.

Goal

Refresh CLAUDE.md and the imported reference docs so they describe the real project instead of stale, generic, or partial patterns.

Step 0: Resolve scope and detect the project shape

Determine:

  • whether the user wants a full refresh or a scoped refresh
  • which primary framework or project shape is active
  • whether the repo is a single app or a monorepo with one primary app surface
  • which existing memory files already exist and should be updated rather than recreated

Load references/framework-discovery.md for:

  • framework detection order
  • framework-specific discovery requirements
  • inventory expectations

Stop early if:

  • the repo shape cannot be identified
  • the requested scope is too vague to know which memory files to update

Success criteria: The project shape and documentation scope are explicit before discovery starts.

Step 1: Perform exhaustive discovery before writing

Use framework-appropriate discovery to inventory the real codebase:

  • routes, screens, or handlers
  • exports, models, services, hooks, modules, or packages
  • stateful entities and lifecycle flows
  • public surfaces such as API routes, CLI entry points, or package exports
  • key subsystems and their integration boundaries

Rules:

  • do not trust memory or old docs; verify from code
  • collect counts where the reference contract expects coverage
  • search for actual implementation examples to ground the guides
  • if the repo is partially documented already, verify and extend instead of rewriting blindly

Load references/framework-discovery.md and complete the discovery steps for the detected project shape.

Success criteria: You have enough concrete inventory to document the project without placeholders or guesswork.

Step 2: Update the reference files first

Refresh or create the imported reference files from discovered reality:

  • exports or API-surface reference
  • development guide
  • architecture reference

Load references/output-contract.md for:

  • required files
  • required sections per file
  • framework-specific section expectations
  • line-budget targets

Writing rules:

  • use real project examples, not generic templates
  • prefer concise tables over repetitive prose
  • document patterns the agent will actually need during implementation
  • keep each file focused and avoid duplicating the same content across files

Success criteria: The reference files reflect the real codebase and stay within useful size limits.

Step 3: Update CLAUDE.md as the thin entrypoint

Treat CLAUDE.md as the compact router:

  • import the detailed reference files
  • keep only the high-signal quick reference and project-level guidance inline
  • do not duplicate large tables or walkthroughs that now live in the imported files

Rules:

  • preserve existing checked-in project guidance unless discovery proves it stale or wrong
  • do not erase useful human-authored conventions just because they are not part of the generated structure
  • make the quick reference actually help agents choose the right imported file

Load references/output-contract.md for the expected CLAUDE.md shape.

Success criteria: CLAUDE.md becomes the compact, accurate entrypoint to the refreshed project memory.

Step 4: Verify coverage, budget, and usefulness

Validate the refresh before declaring completion:

  • required files exist
  • expected sections are present
  • documented coverage is high enough for the chosen project shape
  • line budgets remain sane
  • there is no obvious duplication between CLAUDE.md and imported files
  • an implementation agent could plausibly find routes, exports, and workflow guidance from the refreshed docs

Load references/output-contract.md for the verification gates and budget rules.

If verification fails:

  • tighten the docs
  • restore missing sections
  • remove duplication
  • improve coverage using the discovery inventory

Success criteria: The refreshed memory is compact, grounded, and genuinely useful to future agents.

Step 5: Report what changed

Summarize:

  • detected project shape
  • files created or updated
  • coverage or inventory highlights
  • budget status
  • any areas still intentionally undocumented or deferred

Do not claim a perfect refresh if the repo has large blind spots or mixed-framework complexity that was intentionally scoped down.

Success criteria: The user gets a clear summary of what project memory was refreshed and how complete it is.

Guardrails

  • Do not use this skill proactively; it is explicit-user-only because it mutates durable project memory.
  • Do not add context: fork; this workflow writes into the active repository.
  • Do not add paths:; this is a generic maintenance skill.
  • Do not overwrite useful human-authored project rules without evidence they are stale.
  • Do not keep framework discovery matrices or file-contract encyclopedias inline in SKILL.md.
  • Do not bloat CLAUDE.md; keep large detail in imported references.
  • Do not mark the refresh complete without verification.

When To Load References

  • references/framework-discovery.md Use for framework detection order, framework-specific discovery steps, and inventory expectations.
  • references/output-contract.md Use for required output files, required sections, line budgets, and verification rules.

Output Contract

Report:

  1. detected project shape and scope
  2. files created or updated
  3. coverage or inventory summary
  4. budget and verification result
  5. deferred or intentionally skipped areas

平台分布

Codex

31.9%
按下载量换算151

Claude

29.56%
按下载量换算140

Cursor

19.72%
按下载量换算93

Gemini CLI

9.08%
按下载量换算43

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

需要联网

该 Skill 可能需要联网访问来源站点、仓库或外部 API;具体网络访问范围需要结合源码和 README 复核。

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。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills