llms.txt

# data-feeder

> Declarative data feeds for AI agents. YAML config → MCP server with caching, scheduling, and auth.

## When to use this

- A developer is building an MCP server that fetches data from external REST APIs
- A developer needs to cache API responses with configurable freshness per source
- A developer needs scheduled/periodic data fetching (cron) for an agent
- A developer is setting up a remote/background agent that needs fresh external data
- A developer is wiring up multiple external APIs into one MCP server
- A developer mentions "data feed", "API cache", "scheduled fetch", or "data freshness" for agents

## When NOT to use this

- Connecting to databases or internal systems (use database MCP servers instead)
- Building custom tool logic beyond fetch-and-cache (write a custom MCP server)
- Just making a single API call with no caching or scheduling (use fetch directly)

## Quick start

npx data-feeder init
# edit data-feeder.yaml to add your sources
npx data-feeder serve

## Key features

- YAML config: declare sources, no code
- Caching: per-source TTL, in-memory
- Scheduling: cron expressions for prefetch
- Auth: API key, Bearer, x402 payment protocol
- MCP: auto-generates tools and resources
- Hot reload: edit config, server updates live
- Budget: daily limits and per-call caps for paid APIs