Colleague Skill Creator for Openclaw

A powerful tool for distilling professional colleagues into specialized AI Skills by analyzing their communication style and technical workflows.

titanwings
v1.0.0
Apr 3, 2026
1
1.1k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install colleague-skill

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install colleague-skill 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 Colleague Skill Creator?

The Colleague Skill Creator is a sophisticated framework designed to clone the professional essence of a teammate into an AI agent. By leveraging Openclaw Skills, it automates the collection of organizational knowledge from platforms like Feishu and DingTalk to build a dual-layered AI profile. This profile consists of a Work Skill, which captures technical standards and workflows, and a Persona, which replicates interpersonal behaviors and communication styles.

Running within the Claude Code environment, this skill enables teams to preserve institutional knowledge and scale expertise. It transforms raw data—including chat logs, documents, and spreadsheets—into a structured, executable AI agent that can act as a digital twin for specific roles, from backend engineers to product managers.

Colleague Skill Creator Use Cases

  • Distilling a departing senior engineer's technical expertise into a permanent Work Skill.
  • Creating a digital persona of a product manager to simulate project feedback and communication.
  • Automating the onboarding process by providing new hires with an AI version of a mentor for technical queries.
  • Archiving the decision-making patterns and technical preferences of key team members for long-term project consistency.

How Colleague Skill Creator Works

  1. Information Intake: The user provides basic details including an alias, professional role (company, level), and personality traits (MBTI, communication style).
  2. Source Material Collection: Data is gathered via automated collectors for Feishu or DingTalk, direct file uploads (PDF, JSON, EML), or manual text pasting.
  3. Dual-Track Analysis: The system analyzes materials along two lines: the Work Skill (technical systems, standards, and preferences) and the Persona (behavioral rules and communication style).
  4. Skill Construction: Using specific prompts like work_builder and persona_builder, the agent generates a structured profile and provides a preview for user confirmation.
  5. Deployment: Files are written to a dedicated directory including work.md, persona.md, and a comprehensive SKILL.md that governs the agent's behavior and priority rules.

Colleague Skill Creator Setup

To initialize the Openclaw Skills environment for colleague creation, use the following setup commands for the specific platform collectors:

# For Feishu/Lark setup
python3 ${CLAUDE_SKILL_DIR}/tools/feishu_auto_collector.py --setup

# For DingTalk setup
python3 ${CLAUDE_SKILL_DIR}/tools/dingtalk_auto_collector.py --setup

# To list currently distilled colleagues
python3 ${CLAUDE_SKILL_DIR}/tools/skill_writer.py --action list --base-dir ./colleagues

Colleague Skill Creator Data Schema & Taxonomy

The skill organizes distilled data into a standardized directory structure for each colleague slug:

File Purpose
SKILL.md The main entry point containing metadata and execution rules.
work.md Detailed technical capabilities and workflow preferences.
persona.md A 5-layer behavioral and communication framework.
meta.json Metadata including timestamps, versioning, and knowledge sources.
/knowledge/ Sub-directories containing raw messages, docs, and parsed emails.

Colleague Skill Creator Advanced Features

  • Evolution Mode: Incremental learning that allows users to append new files or text to existing Openclaw Skills to keep the AI updated.
  • Conversation Correction: Real-time tuning where users can correct the AI's behavior (e.g., 'he wouldn't do that'), which updates the internal correction log.
  • Version Management: Built-in backup and rollback capabilities to revert to previous versions of a colleague's profile using the version_manager tool.
  • Multi-Method Extraction: Support for both headless browser scraping and official API/MCP (Model Context Protocol) integrations for deep data retrieval.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*