Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
nanoMuse provides a framework for personal agents to interact with real-world location data. By integrating the Amap MCP server, it enables AI agents to calculate travel times and determine routes via car, transit, bike, or foot within Chinese cities.
This capability allows automation builders to implement precise spatial awareness, helping agents identify nearby points of interest and retrieve current weather conditions for specific urban locations.
WHO IT'S FOR
AI agent / automation builders
integrating cross-platform GUI and API capabilities
MCP server developers
implementing location and weather services in China
Enterprise productivity engineers
automating Lark or Feishu communication workflows
AI agent power users
performing complex multi-option comparisons and decision making
Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.
Follow the documented setup, then try a first task.
nanoMuse for Android
Installs: nanoMuse Android application · nanoMuse
Before you start
An email address or mainland-China phone number for sign-in.
Platform: Android 8.0+, arm64
Download and install https://github.com/nano-muse/nanoMuse/releases/download/v1.0.1/nanoMuse-1.0.1-arm64.apk. Releases use the same signing key and install over the previous version.
Open the app and sign in with an email address or mainland-China phone number to access a model.
An email address or mainland-China phone number for sign-in.
Platform: macOS 12+, Apple Silicon or Intel
Download the DMG matching your Mac: Apple Silicon at https://github.com/nano-muse/nanoMuse/releases/download/v1.0.1/nanoMuse-Desktop-1.0.1-mac-arm64.dmg or Intel at https://github.com/nano-muse/nanoMuse/releases/download/v1.0.1/nanoMuse-Desktop-1.0.1-mac-x64.dmg. Install the app; it is not notarised, so right-click → Open the first time.
Sign in with an email address or mainland-China phone number to access a model.
Intended use, budget, region and relevant constraints.
Access to web_search and web_fetch for factual research.
Make sure compare-options is installed in your agent before trying this task.
Use the compare-options skill to compare the two options I provide for my situation, budget and priorities. Create a comparison page with adjustable criterion weights, sourced facts and a short recommendation explaining what would change your pick.
Meeting name and enough context to identify it, including attendees if needed.
Available calendar, contact, mail, memory, file or goal context; the skill supports partial connector availability.
A workspace for the Markdown brief.
Make sure meeting-prep is installed in your agent before trying this task.
Use the meeting-prep skill to prepare a one-page brief for the meeting I name. Include the attendees, where things stand, open questions, two or three desired outcomes and relevant files to read. Use only available sources, omit unsupported sections and give me the gist in chat.
Any existing inbox rules or reply-style preferences.
Make sure inbox-triage is installed in your agent before trying this task.
Use the inbox-triage skill to sort my recent mail into Reply needed, Waiting on them, To do, Read later and Ignore. Draft short replies for threads that need them, save the drafts for review and highlight the two or three things that should not wait. Do not send messages or create reminders or events.
Claude
Gemini CLI
Amp
DeepSeek