Fork an AI team — Use → make yours → Fork → list
Fork AI team and Fork Grok Bot searchers: on BotShelf, Fork means you start from a ready-to-use team, adapt the blanks and stop rules to your work, then submit your version for review. Public concept page — the interactive form lives on a separate URL.
The loop in four steps
1. Use — open a team on the shelf, fill blanks, paste into Grok Bot Description (or Claude CLAUDE.md / ChatGPT Instructions). Run the job with human Yes before irreversible actions.
2. Make yours — change tone, niche language, instruments, or stop points until it fits how you actually work. Keep one clear job.
3. Fork — from a team page, open Fork to propose your version. Sending the form does not publish by itself; Bot Shelf reviews first.
4. List — after review, your fork can sit on the shelf for others. Paid forks follow the same 80/20 USDT TRC20 rules when approved.
Why Fork instead of copying silently
Forking keeps lineage visible (parent team → your version) and makes it easier for buyers to see what changed. It also keeps the marketplace honest: improvements stay attributable, and review still checks evidence and stop rules.
The interactive fork.html form may be noindex because it expects a parent team context. This SEO page explains the concept and points you to public browse / register / sell paths first.
Good Fork candidates
- Grok Bot teams — adapt desks for your book or inbox
- Fundraising specialists — vendor / SaaS / recurring language for your stack
- Markets desks — gold / FX / index session maps (analysis only)
- Claude Code teams · ChatGPT teams
Browse Grok Bot teams Browse AI teams Sell / list Register
Related
FAQ
Is Fork the same as Git fork?
Conceptually similar — start from a parent, change it, share your version — but here the “repo” is a paste-ready AI team listing, not source code.
Do I need login fluff to understand Fork?
No. Browse public teams, read this page, register when you want free bodies or to submit a Fork/Sell form.
Can I Fork a Grok Bot team for Claude?
Often yes when the job shape is Multi-AI. Rewrite paste instructions for the target runtime and keep stop rules.