Version: 1.0.0
Purpose: Query endpoints split from endpoints.md for context window efficiency
This directory contains query endpoints organized by entity type. This allows AI agents to load only the endpoints they need, reducing context window usage.
Use Case: Load specific entity query endpoints when working with that entity, instead of loading the entire endpoints.md (1153 lines).
Queries are reads. They do not spend charge the way transactions do, but they still hit the node and they still return string numerics — see integration notes before you parse amounts as JSON numbers. If you need a live, sub-second view of the same entities, that is Guild Stack / PostgreSQL, not these REST routes.
Every list endpoint uses Cosmos key pagination: pagination.limit defaults to 100 and is silently clamped to 1000; pagination.count_total is silently forced off (total is always 0); pagination.offset above 1000 is refused (pagination offset exceeds maximum; use pagination key). Walk with next_key. See patterns/pagination.md.
To get Player endpoints:
api/endpoints.md (1153 lines)To get Player endpoints:
api/queries/player.md (~50 lines){
"load": "api/queries/player.md"
}
{
"load": [
"api/queries/player.md",
"api/queries/planet.md"
]
}
Last Updated: January 2025