Runner cache server is now reachable from job containers — root cause was Docker bridge isolation: jobs ran on docker0 while the act_runner cache server sits on gitea-ci_default. Fixed on the runner by setting container.network: gitea-ci_default in its config.yaml (verified: a container on that network gets HTTP 404 from the cache server = reachable; on docker0 it was HTTP 000 = timeout).
Re-adding the two actions/cache steps (opencode + Playwright/npm) reverted in #107. Still continue-on-error as a safety net.
Runner cache server is now reachable from job containers — root cause was Docker bridge isolation: jobs ran on `docker0` while the `act_runner` cache server sits on `gitea-ci_default`. Fixed on the runner by setting `container.network: gitea-ci_default` in its `config.yaml` (verified: a container on that network gets HTTP 404 from the cache server = reachable; on `docker0` it was HTTP 000 = timeout).
Re-adding the two `actions/cache` steps (opencode + Playwright/npm) reverted in #107. Still `continue-on-error` as a safety net.
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Runner cache server is now reachable from job containers — root cause was Docker bridge isolation: jobs ran on
docker0while theact_runnercache server sits ongitea-ci_default. Fixed on the runner by settingcontainer.network: gitea-ci_defaultin itsconfig.yaml(verified: a container on that network gets HTTP 404 from the cache server = reachable; ondocker0it was HTTP 000 = timeout).Re-adding the two
actions/cachesteps (opencode + Playwright/npm) reverted in #107. Stillcontinue-on-erroras a safety net.