dot-skill

Generates persona and work-capability Skills from chat logs, documents, and research sources.

Updated Aug 15, 2026
One-click install
npx skills add https://github.com/jacksonlee-tw/mystock-vue --skill dot-skill-jacksonlee-tw
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: dot-skill
Source: https://github.com/jacksonlee-tw/mystock-vue/tree/main/colleague-skill
Command: npx skills add https://github.com/jacksonlee-tw/mystock-vue --skill dot-skill-jacksonlee-tw

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, pypinyin, playwright, slack-sdk, python-docx, openpyxl, and includes references (resource) components.

What problem does it solve? When a colleague leaves, a mentor moves on, or you want to preserve how someone thinks and speaks, their knowledge and communication style disappear with them. dot-skill distills any person — coworkers, close relationships, or public figures — into a reusable AI Skill that reasons in their frame and responds in their voice. ## Core Features & Use Cases - Three character families: colleague (Work Skill + Persona dual layer), relationship (expression DNA, conflict and repair patterns), and celebrity (six-dimension research pipeline with quality checks). - Multi-source collection: auto-collectors for Feishu, DingTalk, and Slack, plus parsers for email (.eml/.mbox), Feishu JSON exports, PDFs, images, and pasted text. - Evolution and versioning: append new materials with incremental merging, apply conversational corrections, and roll back to any archived version. - Use Case: Provide a departing colleague's name, let the Feishu auto-collector pull their messages and documents, and receive an installable Skill that answers technical questions with their methods and tone. ## Quick Start Ask your agent to install the dot-skill skill from the GitHub repository, then type /dot-skill and follow the prompts to choose a character family and provide source materials.

Frequently Asked Questions about dot-skill

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create an AI skill from a colleague's chat history?▼

Run /dot-skill, choose the colleague family, enter a name and optional profile, then pick a data source. The Feishu, DingTalk, or Slack auto-collectors pull messages and documents, and the pipeline generates a combined Work Skill plus Persona Skill.

What data sources does dot-skill support for skill generation?▼

It supports Feishu, DingTalk, and Slack auto-collection, WeChat chat history via SQLite export, email .eml/.mbox files, Feishu JSON exports, PDFs, images, and pasted Markdown or text. Sources can be mixed or skipped entirely.

Which AI hosts are compatible with dot-skill?▼

dot-skill runs on Claude Code, Hermes, OpenClaw, and Codex. Each host has a dedicated installer script under tools/, and generated character skills can be installed into any of these hosts with one command.

Can I update a generated skill with new materials later?▼

Yes. Append new files or paste new text and the merger prompt analyzes only the delta, merging it into work.md or persona.md without overwriting existing conclusions. Every update is archived, and version_manager.py supports rollback to previous versions.

Why does Slack collection only return 90 days of messages?▼

Free Slack workspaces limit message history access to 90 days; paid plans remove this restriction. The bot also needs channels:history and groups:history scopes and must be invited into each target channel before collection.

What is the difference between budget-friendly and budget-unfriendly celebrity research?▼

Budget-friendly uses three merged research files covering six dimensions with lighter thresholds. Budget-unfriendly requires six separate track files, a research audit, synthesis with triple-gate mental model validation, and a final validation pass including known-answer and voice checks.