A professional AI agent skill to programmatically fetch and aggregate running data from COROS sports watches via official API integration.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install coros-data-skill
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install coros-data-skill using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
The COROS Data Skill is a technical integration designed to connect AI agents with COROS fitness ecosystems. It allows for the automated retrieval of sports activity records, specifically focusing on various running modes. By utilizing this skill within the Openclaw Skills framework, developers can build applications that monitor physical performance, generate training reports, and sync exercise data with other productivity tools.
To get started with this skill for Openclaw Skills, configure your environment variables in the scripts/.env file:
COROS_ACCOUNT=<your_coros_account_email_or_phone>
COROS_PASSWORD=<md5_hashed_password>
You can generate the required MD5 hash using the provided utility:
# Use the genHashedPassword utility to convert your plain text password
import { genHashedPassword } from "./util.js";
const hashed = genHashedPassword("your_password");
console.log(hashed);
The skill processes sports activity data with the following structure:
| Attribute | Type | Description |
|---|---|---|
date |
string | Activity date in YYYYMMDD format |
distance |
number | Total distance covered in meters |
mode |
number | Activity type code (100: Outdoor Run, 101: Indoor Run, 102: Trail Run, 103: Treadmill) |
accessToken |
string | Session token used for authenticated requests |
Loading
An intelligent decision-support tool for evaluating Qunar travel bookings by weighing public trade-offs, pricing, and flexibility.

A dynamic model routing tool that allows users to select specific Claude models for their Telegram-integrated AI agent.

A high-performance search primitive for AI agents to retrieve, filter, and synthesize real-time content from Twitter and global news sources.

A high-agency execution framework that drives OpenClaw agents to push past shallow results and achieve evidence-backed completion.

A local-first habit and goal tracking system designed to provide visual progress, streak maintenance, and accountability for AI agents.

Focus Master is a comprehensive time management CLI tool that combines Pomodoro timers with detailed task tracking and efficiency reporting.








































