# Self-Healing CI: Auto-Retrying "Could not fetch specs from rubygems.org"

> Bundler::HTTPError fetching specs from rubygems.org is a transport failure that survives Bundler's own retries. See the manual fix and how self-healing CI retries the install.

Source: https://latchkey.dev/learn/self-healing-ci/self-healing-bundler-could-not-fetch-specs  
Updated: 2026-09-02

Bundler reporting "Could not fetch specs" has already retried internally four times: the Gemfile is fine and the specs are there, the transport was not.

## 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

### Bundler already retried four times. Why would another retry help?

Bundler's retries happen back to back inside one command, so they all land inside the same bad window. A step-level retry waits seconds between attempts, which is usually longer than the blip.

---

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