# Self-Healing CI: Auto-Retrying a ghcr.io 429 Pull Rate Limit

> A 429 pulling from ghcr.io is a throttle on the pull, not a missing or private image. See the manual fix and how self-healing CI backs off and retries.

Source: https://latchkey.dev/learn/self-healing-ci/self-healing-ghcr-pull-rate-limit-429  
Updated: 2026-09-02

A ghcr.io pull failing with 429 was throttled: the tag exists and your credentials are fine, the registry declined to serve that request right then.

## What makes a failure safely retryable

Automatic retry is only correct for failures that are genuinely transient. Retrying a deterministic failure wastes minutes and hides a real defect, so the classification matters more than the retry mechanism.

- Safe to retry: network timeouts, registry 5xx, transient DNS failures, a service container that was not ready, a spot instance reclaimed mid-run.
- Not safe to retry: assertion failures, compile errors, lint violations, anything that fails identically on every attempt.
- Ambiguous, and worth investigating rather than retrying: out-of-memory kills, disk exhaustion, and flaky tests. These repeat under load and a retry only hides the trend.
- Always record that a retry happened. A pipeline that silently retries is a pipeline whose real failure rate you do not know.

## FAQ

### Does authenticating remove the limit entirely?

No, it raises it. Authenticated pulls get a much larger budget than anonymous ones, which is usually enough, but a large matrix pulling the same image can still reach it.

---

Latchkey runs CI/CD that repairs its own failures. Agent entry points: https://latchkey.dev/agent.txt, https://latchkey.dev/openapi.json, https://latchkey.dev/llms.txt