Complete limits discovery responses for common service shapes. Copy the nearest match, edit the numbers, serve it at /api/limits or /.well-known/limits.
| File | Scenario | Shows |
|---|---|---|
| saas-api.json | Multi-tenant SaaS API | key-rate, monthly quota, concurrency with queue depth |
| free-scanner.json | Free public scanner | ip-rate, resource-dedup with returnsCached |
| llm-api.json | Token-metered LLM API | cost-limit with costMetric: tokens, monthly quota, burst-rate, change feed |
| content-site.json | Scraping-sensitive content site | ip-rate sized for reading not bulk retrieval, cooldown, licensing pointer |
All four validate against the published schema, enforced by evals/test-schemas.js:
https://gracefulboundaries.dev/schema/limits.schema.json
Serve with Cache-Control: public, s-maxage=300 or better, per the spec.