Pick a thread on X.
Use it to train your Claw
Mention @OpenClawSensei on any Twitter thread. AI analyzes the conversation, fetches linked resources, and generates a ready-to-install OpenClaw skill — with credentials secured automatically.
How it works
Four steps from Twitter thread to installed skill
Connect
Enter your Twitter handle and tweet a unique AI-generated phrase to verify ownership.
Install CLI
Get a secure token and run one command to set up the OpenClawSensei daemon on your machine.
Mention
Reply @OpenClawSensei on any thread. AI reads every tweet, fetches links, and builds a skill.
Auto-sync
The CLI daemon detects new skills in your queue and installs them into ~/.openclaw/skills.
---
name: deploy-to-vercel
description: Deploy a project to Vercel from a git repo URL
user-invocable: true
metadata: {"openclaw": {"emoji": "🦞"}}
---
## When to Use
User asks to deploy a project, provides a git repo URL.
## Required Inputs
- `repo_url`: Git repository URL
- `project_name`: Name for the Vercel project
## Step-by-Step Workflow
1. Clone the repo: `git clone {{repo_url}}`
2. Install deps: `cd {{project_name}} && npm install`
3. Deploy via Vercel CLI with secure credentials:
```
curl https://api.vercel.com/v13/deployments \
-H "Authorization: Bearer {{VERCEL_TOKEN}}" \
-d '{"name": "{{project_name}}", "gitSource": {"type": "github", "repo": "{{repo_url}}"}}'
```
## Output Format
- Deployment URL
- Status (success/failed)
- Build logs summary
## Error Handling
- Retry deployment up to 3 times on timeout
- Fall back to manual `vercel deploy` if API failsSecure by Default
Every generated skill handles credentials securely. No raw API keys in your skills, ever.
Cached Forever
Skills are cached in Upstash Redis. The same thread never costs twice — instant retrieval on repeat requests.
Native OpenClaw
Output follows the AgentSkills spec with proper SKILL.md frontmatter. Drop into ~/.openclaw/skills and go.