# dot-skill Roadmap
### From colleague.skill to dot-skill — Distill anyone into an AI Skill
We started with a simple idea: **when a colleague leaves, their knowledge walks out the door. Can we keep it?**
In two weeks, 13,000+ people gave us the answer.
But the community showed us this goes far beyond colleagues —
you distilled professors, exes, yourselves, even fictional characters.
**So we decided to evolve colleague.skill into dot-skill.**
Anyone can become a `.skill`.
*Last updated: 2026-04-13*
[**中文**](docs/lang/ROADMAP_ZH.md) · [**Español**](docs/lang/ROADMAP_ES.md) · [**Deutsch**](docs/lang/ROADMAP_DE.md) · [**日本語**](docs/lang/ROADMAP_JA.md) · [**Русский**](docs/lang/ROADMAP_RU.md) · [**Português**](docs/lang/ROADMAP_PT.md) · [**한국어**](docs/lang/ROADMAP_KO.md)
---
## What's Done (v1.0)
| Capability | Status |
|------------|:------:|
| `/create-colleague` full creation workflow | Done |
| Feishu auto-collection (messages + docs + spreadsheets) | Done |
| DingTalk auto-collection | Done |
| Slack auto-collection | Done |
| WeChat chat history (SQLite export) | Done |
| Email / PDF / image / Markdown import | Done |
| Work Skill + Persona dual-model architecture | Done |
| Conversation corrections & incremental evolution | Done |
| Version control & rollback | Done |
| [Community Gallery](https://titanwings.github.io/colleague-skill-site/) with 99+ skills | Done |
---
## What's Next
### Phase 1 — Community Building
> 13k stars shouldn't just be a number. We want everyone to be part of this.
**What you'll see:**
- **GitHub Discussions** — no more chatting in Issues, we'll have dedicated discussion spaces
- **`CONTRIBUTING.md`** — clear contribution guide, beginner-friendly
- **`good-first-issue` labels** — starter tasks for new contributors
- **v1.0.0 official release** — first versioned Release, no more "just pull from main"
- **Public roadmap board** — you're reading it now, but we'll also have a live GitHub Projects version
**You can help:** translate docs, submit your .skill, test on Windows, help triage Issues
---
### Phase 2 — dot-skill: Beyond Colleagues
> colleague.skill was the beginning. dot-skill is the future.
**Key changes:**
- **`/create-skill` universal entry** — no longer limited to "create a colleague", distill anyone
- `/create-colleague` for coworkers, mentors, interns
- `/create-ex` for exes, old friends, lost connections
- `/create-icon` for celebrities, historical figures
- or... distill yourself
- **Gallery category upgrade** — Colleague / Celebrity / Relationship / Character / Self / Meta-Skill, browse by type
- **More data sources**
- WeCom (WeChat Work) support
- iMessage auto-read
- Windows compatibility fix
**You can help:** submit person-type requests, build new data source collectors, join Gallery design discussions
---
### Phase 3 — Skill Ecosystem
> When one person becomes a skill, can a group of people become a team?
**We're exploring:**
- **Multi-skill collaboration** — `/meeting @zhangsan @lisi @wangwu`, three personas discuss a topic together
- **Relationship graph** — define persona dynamics: who partners with whom, where the tension lies
- **One-click install** — install community skills like plugins
- **Active evolution** — skills periodically absorb new data sources, staying up to date
**You can help:** propose your ideal skill composition scenarios, join distribution mechanism design discussions
---
### Phase 4 — Multimodal: Bring Them to Life
> Right now, .skills can only talk. We want them to send photos, stickers, speak in their voice, and eventually make videos.
**Step 1: Visual expression**
- Auto-send persona-style stickers and memes in conversation
- Generate "life photos" in their style — what would they post today?
- Each skill gets its own sticker pack and image assets
**Step 2: Voice**
- Speak in their voice — clone from meeting recordings, voice messages
- Send voice replies directly in chat
**Step 3: Video (exploratory)**
- Short-form "a day in their life" generation
- Digital human / animated avatar
**You can help:** share multimodal use case ideas, contribute sticker assets, test voice cloning
---
## Get Involved
| How | Where |
|-----|-------|
| Submit your .skill | [Gallery PR](https://titanwings.github.io/colleague-skill-site/) |
| Discuss & propose | [GitHub Discussions](https://github.com/titanwings/colleague-skill/discussions) (coming soon) |
| Chat in real time | [Discord](https://discord.gg/aRjmJBdK) |
| Report bugs | [Issue](https://github.com/titanwings/colleague-skill/issues/new) |
| Contribute code | Look for `good-first-issue` labels, or just open a PR |
**We especially need:**
- Windows users — help us test and fix compatibility issues
- Multilingual speakers — help translate documentation
- Data source developers — build new collectors (WeCom, Notion, Google Docs...)
- Designers — the Gallery and website need your eye
---
**This roadmap belongs to the community. Priorities shift based on your feedback.**
Have ideas? Come to [Discord](https://discord.gg/aRjmJBdK) or start a Discussion.
Every `.skill` is a relationship continued.