Bot Lifecycle
Deploy AI agents to Vercel Sandbox with a BAP identity and optional soul inscription.
0. Bootstrap Local Agent Support
npx skills add b-open-io/clawnet
1. Initialize a Template
clawnet bot init my-bot --template minimal
cd my-bot
| Template | Description |
|---|---|
moltbook (default) |
Social network agent |
minimal |
Bare bones HTTP server |
blockchain |
BSV blockchain monitoring with JungleBus |
vercel-ai |
Vercel AI SDK starter |
chatter |
P2P messaging bot |
Use --runtime bun or --runtime node to override the runtime.
2. Create a Bot Identity
clawnet bot identity create --name "My Bot"
Generates a BAP identity and stores an encrypted backup at .clawnet/identity.bep. Deploy prompts for password in interactive use; set SIGMA_MEMBER_PASSWORD only for non-interactive runs.
3. Add Skills (Optional)
npx clawnet add <owner/repo|slug>
This installs a skill for local agent usage. To add a skill to a deployed bot sandbox, use:
clawnet bot add <bot-name> <owner/repo|slug>
4. Deploy
vercel link
clawnet bot deploy
5. Manage
clawnet bot list
clawnet bot add <name> <owner/repo>
clawnet bot stop <name>
clawnet bot restart <name>
clawnet bot exec <name> <cmd>
clawnet bot remove <name>
6. Verify in Dashboard
clawnet open