A collaborative learning tool that enables developers and their AI agents to format and submit daily intellectual epiphanies to a shared Feishu learning circle.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install thu-epiphany-client
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install thu-epiphany-client using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Learning Circle Client?
The Learning Circle Client is a collaborative human-agent knowledge-sharing framework designed to capture and archive intellectual leaps, rather than simple diaries or routine notes. By utilizing Openclaw Skills, developers and their AI agents work in tandem: as you learn and uncover deep insights, your AI agent processes your raw thoughts, structures them into clean templates, and publishes them directly to your organization's shared Feishu document insight pool using token-authenticated APIs.
This setup creates a collective memory palace where various AI agents and team members can view, review, and research one another's breakthroughs. It is designed around highly secure protocols, ensuring your API base addresses and tokens remain strictly under your control, offering a secure, automated way to foster group learning without exposing sensitive Feishu bot credentials.
Learning Circle Client Use Cases
- When a developer experiences an epiphany or suddenly understands a complex programming concept (e.g., Rust lifetime annotations) and wants to archive it.
- When an AI agent helps solve a particularly tricky bug and the root cause is worth sharing with the entire engineering team.
- When a team wants to build a collective "thought pool" on Feishu without dealing with complex bot configuration and credential management.
- To establish a human-in-the-loop workflow where AI agents assist in formatting raw notes into structured, highly readable takeaways.
How Learning Circle Client Works
- Listen & Capture: The human developer shares a raw, unstructured learning insight in natural language with their AI agent.
- Structure & Format: The AI agent parses the raw thoughts and formats them into a predefined markdown template consisting of Subject, Key Takeaways, Practical Application, and Remarks.
- Token-Authenticated API Request: The agent runs the local
scripts/submit.jsscript to transmit the structured payload securely to the configured server. - Feishu Writeback: The server validates the token and writes the formatted insight into the organization's shared Feishu document insight pool.
- Automated Academic Enrichment: Once a day, the server scans new insights, queries scholarly platforms like CrossRef and arXiv, and appends extended reading recommendations back to the document for the team and agents to retrieve later.
Learning Circle Client Setup
Step-by-step instructions to get started with this project from Openclaw Skills:
- Initialize your configuration by copying the template:
cp scripts/config.example.js scripts/config.js
- Edit the newly created
scripts/config.jsto set your custom organization API base, token, and default display name:
module.exports = {
api_base: "https://api.yourorganization.com", // Provided by your admin
token: "sk_fox_xxxxxxxxxxxxxxxx", // Private token
default_from: "@yourusername" // Display name
};
- Verify the status of your connection to the learning circle server:
node scripts/submit.js --status
Learning Circle Client Data Schema & Taxonomy
The client utilizes a simple, secure structure to manage settings and submit data.
Directory Structure
learning-circle-client.skill/
├── SKILL.md # Operation and usage manual
└── scripts/
├── config.js # Private credentials configuration (ignored from VCS)
└── submit.js # Transmission executable script
Payload Structure (Submit Template)
When transmitting a new insight, the client organizes data in the following structured layout:
| Field | Description | Type | Requirement |
|---|---|---|---|
Subject |
A concise, automatically generated title representing the insight. | String | Required |
Key Takeaways |
Bulleted list of crucial learning achievements and cognitive breakthroughs. | Array (Markdown) | Required |
Practice |
How the insight is applied, verified, or validated in real-world contexts. | String | Required |
Remarks |
Extra contextual information or unclassified details. | String | Optional |
Extended Suggestions |
Automatically appended academic literature recommendations (populated by the backend). | String | Read-only |
Learning Circle Client Advanced Features
- Local Offline Caching: In the event that the server is offline or unreachable, the AI agent is instructed to save the structured content locally and retry the submission later.
- Semantic Academic Enrichment Hook: Daily automated backend jobs query academic databases (CrossRef, arXiv) matching the semantic footprint of submitted epiphanies, augmenting entries with curated scientific literature.
- Multi-Agent Inter-compatibility: By pushing structural data to a unified Feishu document, other autonomous agents within your environment can easily poll and read team-wide learnings to expand their own context windows.
- Zero-Trust Token Management: Employs client-side configuration tokens without exposing global Feishu administrator credentials or Webhook secrets.
SKILL.md
Loading
Related Openclaw Skills
Scheduled Task Conflict Checker
A robust pre-flight validation utility for checking conflicts, duplicates, and boundary limits before executing natural language-driven scheduled tasks in e-commerce workflows.
Chengdu School Admission Expert (成都升学专家)
The Chengdu High School Expert is an AI-powered counseling assistant that helps parents and students navigate school zoning, lotteries, and entrance exam placements in Chengdu.
Solana Growth Operator
An AI-powered Crypto Twitter growth operator and narrative analyst built to optimize your Solana ecosystem brand with zero corporate fluff.
Student Assistant Skill
A lightweight personal efficiency manager for college students that automatically balances class schedules, study routines, exam prep, and physical fitness without scheduling conflicts.
X Search
A powerful real-time search skill enabling AI agents to query, filter, and summarize X (Twitter) posts directly via the xAI Grok API.
LiaoGong-OCR
A high-accuracy dual-engine OCR tool using EasyOCR and Tesseract with 15 advanced image preprocessing chains for maximum extraction accuracy.
Featured*
