titanwings

你们搞大模型的就是码奸,你们已经害死前端兄弟了,还要害死后端兄弟,测试兄弟,运维兄弟,害死网安兄弟,害死ic兄弟,最后害死自己害死全人类

437
15
100% credibility
Found Mar 30, 2026 at 437 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

This repository provides a tool to generate AI personas that emulate former colleagues' work styles and personalities by processing their messages, documents, and communications from various sources.

How It Works

1
😢 A colleague leaves

Your teammate quits or moves on, leaving a gap in knowledge and teamwork.

2
🔧 Add the tool to your AI helper

Place this handy kit into your AI assistant's skills area so it's ready to use.

3
Start making a digital twin

Type a simple command, describe their job, company style, and personality traits like 'careful perfectionist'.

4
Share their past work
🤖
Auto-grab from team chats

It pulls messages and docs straight from apps like Feishu or DingTalk by name.

📎
Upload files yourself

Add emails, docs, screenshots, or pasted text easily.

5
🧠 It learns their ways

The tool studies everything to capture how they work, decide, and talk.

🎉 Chat with your colleague again

Call them by name in your AI, get advice in their exact style, and keep projects moving smoothly.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 437 to 437 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is colleague-skill?

colleague-skill is a Python tool that ingests a colleague's chat logs, docs, emails, and screenshots from Feishu, DingTalk, or PDFs to generate AI skills mimicking their expertise and style. Run `/create-colleague` in Claude Code or OpenClaw, add a description like "ByteDance L2 backend engineer, INTJ blame-shifter," pick sources, and get a slash-invocable skill split into work knowledge (workflows, standards) and persona (decision rules, voice). It revives lost context from quitters, interns, or transfers as "cyber-immortality" for your AI agent.

Why is it gaining traction?

Auto-pulls data from Feishu/DingTalk APIs or browser (no manual exports), supports tags for company cultures like ByteDance-style or Huawei 13-21 levels, and evolves skills via appends, corrections, or rollbacks without overwriting. Users get instant `/zhangsan` commands for code reviews in that exact passive-aggressive tone, plus work-only or persona-only modes. 437 stars reflect appeal in turnover-heavy teams tired of knowledge silos.

Who should use this?

Backend engineers at ByteDance, Alibaba, or Tencent reconstructing ex-colleague APIs from Feishu group chats. QA/ops folks simulating intern handoffs or mentor decisions from DingTalk docs/emails. Teams with high churn needing "ByteDance L2-1 perfectionist" skills for daily blame games or N+1 fixes.

Verdict

Grab it if Feishu/DingTalk dominates your stack—quick wins for colleague skills despite beta bugs. 1.0% credibility and 437 stars signal early promise but unproven scale; docs are solid, test on non-prod first.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.