# GRAPHQL Performance

> Improve GraphQL performance with DataLoader batching, demand control, compression, persisted queries, and operation-level tracing. Use when fixing N+1 resolvers or slow GraphQL operations.

- Skill: `deangrant/graphql-performance` (Agent Skill, multi-file: 4 files)
- Install (CLI): `npx skillmds@latest add deangrant/graphql-performance`
- Raw SKILL.md: https://api.skillmd.com/api/skills/deangrant/graphql-performance/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Integrations & APIs
- Author: deangrant (https://skillmd.com/u/deangrant)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/deangrant/graphql-performance

---


# GraphQL Performance

Use this skill when GraphQL is **slow or chatty** under load.

---

## 1. Client

Normalized store + global IDs to avoid redundant fetches.

---

## 2. Transport

POST default; GET for cacheable queries when supported. Persisted/trusted
queries when URLs are long. Enable gzip/brotli.

---

## 3. Resolvers

**DataLoader** (or equivalent) per entity to batch N+1. Per-request loader
cache. Domain batch methods in the business layer. Consider selection-set → SQL
when ORM allows.

---

## 4. Demand control + observability

Paginate; max `first`; depth/breadth/complexity limits; gateway rate limits.
Trace by `operationName` and field path; OTel; latency/error metrics.

---

## 5. Quick checklist

- [ ] No 1-query-per-list-item resolvers.
- [ ] Loaders scoped per request.
- [ ] Caps on depth/`first`/complexity.
- [ ] Field-level traces for slow paths.

See [reference.md](reference.md) and [examples.md](examples.md).