Diagnose failing CircleCI pipelines and apply minimal root-cause fixes.
npx skills add https://github.com/openai/plugins --skill circleci-builds
Please help me install this Agent Skill. Skill: circleci-builds Source: https://github.com/openai/plugins/tree/main/plugins/circleci/skills/builds Command: npx skills add https://github.com/openai/plugins --skill circleci-builds
Failing CircleCI pipelines are time-consuming to triage, and it is easy to mask deterministic failures with blanket retries or apply broad fixes that introduce new risk. This Skill turns failing builds into actionable fixes by isolating the first failing signal, classifying the failure type, and applying the smallest safe patch.
Ask the assistant to investigate why my CircleCI pipeline failed on the latest commit and propose a minimal fix with a validation plan.
High-intent search queries and answers about installing and using this skill.
Start by recording the first failing job and step rather than every downstream failure, then classify the issue as config, environment, dependency, test regression, or external. Gather the pipeline identifier, branch, commit SHA, and key log lines before proposing any fix.
Rerun the same commit and check whether it passes without any code or config change. Network timeouts, executor interruptions, and tests with a history of passing on rerun suggest transient causes, while repeatable assertion failures and config validation errors are deterministic.
Use automatic reruns only after classifying the failure as plausibly transient, and scope them to the smallest failing unit such as a single step. For flaky tests, prefer circleci tests run with rerun failed tests once JUnit metadata is configured, rather than rerunning entire workflows.
Repeated failures on the same commit after reruns indicate a deterministic defect, not a transient issue. Stop escalating retries and fix the root cause in configuration, dependency versions, environment setup, or test code.
Retries only mitigate temporary failure modes like network timeouts or infrastructure instability and cannot fix deterministic defects. Blanket workflow reruns without evidence of transient behavior mask real bugs and waste compute, so retry counts should stay low and targeted.
muiClassify CircleCI workflow failures by root cause into flakes, outages, or bugs.
buisihung11Monitors Nx Cloud CI pipeline executions and orchestrates automated fixes for common failures.
rwc9uDiagnose CircleCI failures using Datadog test visibility data.
micaelmaltaConfigure CI/CD pipelines across GitHub Actions, GitLab CI, Jenkins, and CircleCI.
