AGENTKIT/WORKS
Free kit

integration · tripwire tier

Xata Database Kit

Turn any agent into a database operator — schema design, CRUD, full-text search, aggregation, one skill file.

✓ Claude CodeCodex CLIHermes Agent30+ tools via the open standard
PRICE$34$24
Buy now — $24

One skill file that routes all structured storage through Xata so your agent stops dumping data into JSON files that break at 10K records. Handles schema design, CRUD operations, full-text search with typo tolerance, aggregation, and free-tier budget tracking. Chains naturally with the Zyte Scraper Kit — scrape data, store it, query it, export it. All in agent language.

$ tree xata-database-kit/✓ actual contents
xata-database-kit/
├── README.md                  ← human overview + quick start
├── INSTALL.md                 ← agent reads this to install itself
├── install.json               ← manifest: destinations + verify checks
├── LICENSE                    ← MIT — commercial use OK
├── SPEC.md                    ← (authoring spec, excluded from zip)
├── skills/
│   └── xata-database/
│       └── SKILL.md           ← the skill: schema, CRUD, search, free tier
├── templates/
│   └── mcp-config.yaml        ← ready-to-paste MCP server block
└── examples/
    └── prompts.md             ← 10 known-good prompts

Why it's different

  • ▸ Works via direct API (Python) or MCP/Composio — your agent picks the right mode
  • ▸ Full-text search with typo tolerance built into the query procedure
  • ▸ Free-tier budget tracking warns before hitting 750K record ceiling
  • ▸ Pipeline-ready: chains with Zyte Scraper Kit (scrape → store → query → export)

Questions, answered

Do I need a paid Xata account?+

No. Xata's free tier gives you 750,000 records, 15GB storage, and 15 branches — enough for years of agent-driven data work. No credit card required.

Does this replace PostgreSQL / Supabase?+

For agent-driven workflows, yes. Xata gives your agent a REST API instead of connection strings, ORM configs, and schema migrations. For production apps with millions of users, keep your Postgres — this is the agent's database, not your app's.

How does this work with the Zyte Scraper Kit?+

They're designed to chain: Zyte scrapes the data, Xata stores it. The kit includes pipeline prompts that go end-to-end — 'scrape books.toscrape.com and store all results in Xata' — one sentence, two kits, done.

RUNS THESE PLAYBOOKS

Run a one-person content pipeline with agentsFor developersFor indie hackersFor agencies