> ## Documentation Index
> Fetch the complete documentation index at: https://docs.early-bird.space/llms.txt
> Use this file to discover all available pages before exploring further.

# Limits

> Why Early Bird keeps request patterns conservative and how to work within those constraints.

# Limits

Early Bird is designed to be conservative by default.

## Core principles

* prefer cached summaries over row scans
* prefer one page before multiple pages
* prefer targeted slices over exhaustive exports

## Why these limits exist

Early Bird currently uses cost-sensitive infrastructure, so agent and API usage should stay disciplined.

## Expected product behavior

* small default page sizes
* explicit pagination
* no public full-table hydration on first request
* clear signaling when results are partial or truncated

## Agent behavior guidelines

* do not crawl every page unless the user explicitly requests exhaustive retrieval
* stop once the question can be answered confidently
* explain when the answer is based on a sample or a capped result set

## Future public policy

The public agent-facing layer is expected to add:

* request budgets
* rate limits
* query-cost metadata
* stronger guardrails around deep scans