TableModel classes. Notebooks can still use pxt.create_table().
Set Up Your AI Tool
- Cursor / Windsurf
- Claude Code
- Any LLM
To install the skill, run
npx skills add. Do not copy this repository’s AGENTS.md into an application; that file is for Pixeltable contributors.What the agent should write
Write an application file, then create its tables withpxt schema update app.py my_app. When you insert a row, computed columns run automatically.
t = pxt.get_table('my_app.articles') and insert rows, or serve the same file over HTTP. To run that file on Pixeltable Cloud, sign in with pxt login and run pxt db update, then pxt schema update, then pxt service update, against pxt://org:main: Pixeltable Cloud.
Choose a Pixeltable MCP server
The Pixeltable plugin packages the Skill for Claude, Cursor, and ChatGPT. Neither the plugin nor the Skill configures an MCP server; add the one the job needs.
For Cloud changes and deployment, use the
pxt CLI after pxt login. The hosted Cloud MCP only reads.
Catalog inspection without importing Python: pxt ls --json, pxt describe --json, pxt errors. The pxt CLI (v0.7.4+) exposes --json on inspection and query commands. See the CLI.
To call remote MCP tools from a table column, use pxt.mcp_udfs(url) (see Agents as tables).
Next
Quickstart
Write
app.py, create the tables, then start the endpoints.Agents as tables
Register tools and run
invoke_tools as computed columns.Starter kit
Templates for serving, batch, and FastAPI + React.
Cloud
The same
app.py. pxt db update, then pxt schema update and pxt service update against pxt://org:main.