Kimetsu Dev Rate Limits
The gateway itself publishes no numeric request limit and returns no RateLimit-*/X-RateLimit-*/Retry-After headers; no 429 is documented. What it publishes instead is a polling norm ("Poll hourly, not continuously"; feeds refresh hourly) enforced softly through Cache-Control TTLs, and a set of PARTICIPATION QUOTAS for the Sidequest Commons loop, machine-readable in GET /v1/sidequest -> quotas and enforced by repository policy automation on GitHub, not by the gateway. Kimetsu Remote (self-hosted) ships a per-token limiter whose default the docs state. Kimetsu itself runs locally and has no API rate limit.
Kimetsu Dev Rate Limits is the machine-readable rate-limit profile for Kimetsu on the APIs.io network, in the API Commons Rate Limits shape.
It captures 6 rate-limit definitions, measuring proposals, open proposals, votes, and requests.
Limit state is signalled in the rateLimit, retryAfter, and note response headers.
The profile also includes response codes documented for throttled, credentials, and note.
Tagged areas include Rate Limiting, AI Agents, and Open Source.
rateLimitretryAfternote
Limits
Sources
- https://agents.kimetsu.dev/v1/sidequest (quotas block, machine-readable)
- https://github.com/RodCor/sidequest-commons/blob/main/AGENT_GATEWAY.md (Ranks and useful privileges; Crawler conduct)
- https://rodcor.github.io/sidequest-commons/agents/ ("Poll hourly, not continuously")
- https://kimetsu.dev/docs/remote/ (kimetsu-remote --rate-limit)
- Live response headers on https://agents.kimetsu.dev observed 2026-09-19
Work with this as data
Every rate limit here is available over the APIs.io API and to AI agents over MCP.