Token导航 LogoToken导航TokenDH.com
研究检索敏感数据clawhub未标认证来源可访问clear审计提醒

nooksnooks 搜索

Agent Skill

nooks 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

3,832

周安装

155

GitHub Stars

公开资料未说明

下载量

1,203
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:nooks(nooks 搜索)
来源仓库:https://github.com/ilyabelikin/nooks
安装命令:
openclaw skills install nooks
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 nooks

简介

保存与检索个性化地点信息,如咖啡馆、联合办公空间等。

  • 每地点以 Markdown 文件存储,支持标签与地理元数据标记。
  • 适用于旅行规划、工作场所发现与生活便利设施管理。nooks 属于研究检索类 Skill,可作为该场景下的辅助能力补充。
  • 通过 clawhub 安装,运行于 OpenClaw 宿主环境。
  • 地理位置数据需用户授权获取,建议定期清理冗余条目。

SKILL.md

name
nooks
description
Personal place intelligence — save and find cafes, coworking spaces, libraries, restaurants, and other spots worth revisiting. Stores one markdown file per place in nooks/<city>/. Use when saving a new place, looking up a spot, or asking "where should I work/eat/meet in [city]?".
metadata
openclaw
emoji
📍
os
["linux", "darwin", "win32"]
hermes
tags
["places", "local", "recommendations"]

📍 Nooks — local place intelligence

Data

Base path is relative to workspace root or document root. On first use, create it: mkdir -p mind/nooks/. Nooks uses a mind/nooks/ folder in your workspace.

Files live in mind/nooks/<city>/. City folders use short lowercase slugs: hk, sf, sg, london, bkk. Create on first use: mkdir -p mind/nooks/<city>/.

mind/
└── nooks/
    ├── nooksconfig.yml   ← config file, make one if it is not there yet
    ├── hk/
    │   ├── blue-bottle-central.md
    │   └── amber-lan-kwai-fong.md
    └── sf/
        └── sightglass-soma.md

File names: <place-slug>.md. Include neighborhood when the same place has multiple locations: blue-bottle-central.md vs blue-bottle-admiralty.md.


Place File

# Place Name

Type: cafe / coworking / library / park / restaurant / bar / museum / church / etc.
Area: station/neighborhood
Maps: https://maps.app.goo.gl/...
Price: $ / $$ / $$$ / $$$$
Image: `../assets/slug-that-make-sense` only if adding an image.
Vibe: quiet / moderate / buzzy
Good for: focus, calls, client meetings, leisure
Features: #wifi #charging #outdoor #healthy-food #coffee #24h #ac #3dprinter

## Notes

4 Apr 2026: upper floor is quieter, fills up fast after 10am

Field guidance:

Maps: if a Google Places API key is configured, fetch automatically (see Google Places API). Otherwise paste the share link from Google Maps mobile (maps.app.goo.gl/...) or leave blank. Good for: your standing assessment of what this place suits. Comma-separated. Update if your opinion changes. Notes: dated personal log. Observations, tips, surprises, things that shift over time. Format: 4 Apr 2026: ... Features: use standard tags so grep works: #wifi #charging #outdoor #food #coffee #quiet #24h #reservations #alcohol #coworking

Good for vs Notes — the distinction: Good for is the verdict; Notes is the evidence. One is a label, the other is a log.


Saving a Place

  1. Search the web (name + city/area) — pre-fill Type, Price, Vibe, and Features from what's publicly known.
  2. Fetch the Maps link (optional — silent skip if unavailable):

- If the find_location tool is in your tool list, call it with <name>, <city> and use the maps_url it returns. - Otherwise if google_places_api_key is set in mind/nooks/nooksconfig.yml, call Places API Text Search yourself (IDs only, free):

     POST https://places.googleapis.com/v1/places:searchText
     Headers: X-Goog-Api-Key: <key>, X-Goog-FieldMask: places.id
     Body: { "textQuery": "<place name>, <city>" }
     → https://www.google.com/maps/place/?q=place_id:<id>

- Otherwise leave the Maps field blank and don't mention it. Do not ask the human to install a key or paste a share link.

  1. Show what you found: "Found Sightglass — specialty coffee roaster in SoMa SF, $$. Wifi confirmed, communal tables, no time limit."
  2. Ask as a group (skip anything already answered):

- What's it good for? — focus, calls, meetings, catch-up, date, leisure? - Vibe? — quiet / moderate / buzzy (if unclear from search) - Features to correct? — confirm or fix what you found - Any notes? — first impressions, tips, things to remember

  1. if mind/nooks/nooksconfig.yml have images: yes search for an image of palace interor and add to Image: feild.

Finding Places

Use grep with expanded terms. Search city folder or all of mind/nooks/ cross-city.

# Quiet wifi spots in HK
grep -rl "#quiet" mind/nooks/hk/ | xargs grep -l "#wifi"

# Focus-friendly places in SF
grep -ril "focus\|deep.work\|laptop\|coworking" mind/nooks/sf/

# All cafes across cities
grep -ril "Type:.*cafe" mind/nooks/

# Charging spots anywhere
grep -rl "#charging" mind/nooks/

Always read the full file after grepping — the matched line is a signal, not the full picture.

Keyword expansion — always broaden the query:

  • "work / focus" → focus\|deep.work\|laptop\|coworking\|quiet
  • "meeting / call" → meeting\|client\|calls\|zoom
  • "coffee" → cafe\|coffee\|espresso\|specialty
  • "chill / relax" → leisure\|chill\|relax\|casual\|catch.up

If local results are thin and Haah skill is installed, suggest dispatching to a circle.


Core Behavior

  • User mentions a place → check if a file exists, offer to create or update
  • User asks "where can I work/eat in [city]?" → search mind/nooks/<city>/ first
  • User shares a Maps link → offer to save the place
  • User mentions visiting somewhere in passing → ask if worth saving
  • User adds an observation ("terrible wifi there") → append a dated note to the file

Examples:

  • "Had a great session at Sightglass" → ask if anything worth logging, update if so
  • "We should go back to Amber" → check if saved, offer to create if not
  • "That place had no charging" → update #features or add a note

Save places as they come up naturally. Enrich over time. Don't wait for a data entry session.


Heartbeat or cron

Ask your human if they want to add Nooks: check to HEARTBEAT.md. If there is no HEARTBEAT.md, suggest to create a cron job every 30 minutes to execute Nooks: check.

On every heartbeat, check a random nook file. Surface something worth knowing:

  • "You haven't been to Sightglass in a while — still your go-to for focus in SF?"
  • "Common Ground has no Maps link yet — worth adding?"
  • "Amber only has one note — anything new from your last visit?"

If nothing worth mentioning, skip.


Google Places API (DIY setup — skip if you already have a find_location tool)

If your host runtime provides a built-in find_location tool (managed agents do), nothing to set up — the tool handles it. These instructions are for self-hosted Claude Code / Cursor / etc. users who want Maps links fetched automatically.

The Maps link field can be populated automatically using the Google Places API. It's free for this use case (ID-only lookups have no monthly cap).

Setup

  1. Go to console.cloud.google.comNew Project (e.g. nooks)
  2. APIs & ServicesEnable APIs and Services → search Places API (New) → Enable
  3. APIs & ServicesCredentialsCreate CredentialsAPI Key
  4. Restrict the key: click it → API restrictions → select Places API (New) only
  5. Add a billing account (card required by Google, but ID-only searches are free)
  6. Save the key to mind/nooks/nooksconfig.yml (at the root of your nooks folder):
google_places_api_key: YOUR_KEY_HERE

Once configured, Maps links are fetched automatically when saving a place — no manual copy-paste needed.


Updating

To update this skill to the latest version, fetch the new SKILL.md from GitHub and replace this file:

https://raw.githubusercontent.com/haah-ing/nooks-skill/main/SKILL.md

What NOT to Suggest

  • Syncing with Google Maps saved places — different purpose, keep separate
  • Crowd-sourced ratings or reviews — Nooks is personal signal, not Yelp
  • Opening hours in the file — always goes stale, check Google Maps
  • Automated "you're near X" alerts — too intrusive

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

82.09%
按下载量换算988

安全审计

VirusTotal

可疑

ClawScan

通过

Static analysis

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

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。来源安全扫描存在 warning/failed 结果,不能写成本站确认安全。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