Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIDeepSeekDeepSeekMuseMuse+22 more

    ci pipeline diagnosis

    1

    - **Metric Computation** generates hard data on failure rates and build reliability without manual spreadsheet work.

    $1.99

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    ci pipeline diagnosis

    ci pipeline diagnosis

    Example session with this skill installed

    Analyze these last 20 pipeline runs from GitHub Actions. Several builds failed during the integration-test step but others passed. I need to know if it is a bottleneck or just flaky.

    • Read your context and instructions
    • Compiled the ci pipeline diagnosis

    The diagnostic script has completed the analysis of 20 pipeline runs. It identified a high failure rate in the integration-test step and clustered three distinct error patterns.

    CI Pipeline Diagnosis — Engineering Team, 2023-10-24

    1. Summary

    • Total Runs: 20
    • Success Rate: 65% (13/20)
    • Failure Rate: 35% (7/20)
    • Bottlenecks Found: integration-test (p90: 450s)
    • Flaky Steps Found: integration-test, lint-check

    2. Step Analysis

    Step NameExecutionsFailuresRateMean Dur.p90 Dur.Flags
    build-assets2000%45s52s
    lint-check20210%12s15sFLAKY
    integration-test20525%310s450sBTLNK, FLAKY

    3. Error Clusters

    • Pattern A (3 occurrences): ETIMEDOUT: connection refused to database:5432. Affected steps: integration-test.
    • Pattern B (2 occurrences): Error: Process completed with exit code 137. Affected steps: integration-test.
    • Pattern C (2 occurrences): Trailing whitespace detected. Affected steps: lint-check.

    4. Findings

    • High-Impact Flakiness: integration-test shows non-deterministic failures related to database timeouts.

    Resource Constraint: Exit code 137 (OOM) suggests the test runner is running out of memory during heavy integration cycles.

    5. Warnings and Data Quality

    • W002: Missing duration data for 1 run in setup-env.

    6. Recommendations and Items for Human Review

    • Increase memory limits for the integration-test runner.
    • Implement a retry mechanism for database connections in the test suite.
    • Management decision required for infrastructure resource allocation.

    Next steps

    1. Review the database connection timeout settings in the integration test config.
    2. Monitor memory usage on CI runners during the integration-test step.
    3. Check lint-check local pre-commit hooks to eliminate simple failures.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    CI pipelines often fail for reasons that aren't immediately obvious, such as transient flakiness, silent bottlenecks, or recurring error patterns across multiple runs. This skill analyzes historical pipeline data to surface exactly where builds are breaking and why.

    What it does

    • Failure Analysis clusters similar error messages across multiple runs to identify root causes instead of individual symptoms.
    • Flakiness Detection identifies steps that fail inconsistently to help isolate non-deterministic tests or environment issues.
    • Bottleneck Identification calculates mean and p90 durations for every step to pinpoint which parts of the CI cycle delay development.
    • Metric Computation generates hard data on failure rates and build reliability without manual spreadsheet work.
    • Prioritized Diagnostics produces a structured report ranking issues by impact with specific recommendations for remediation.

    How it works

    1. Provide pipeline run data, such as exported logs or CI platform JSON exports containing step names, statuses, and durations.
    2. The agent extracts this data into a structured format for the internal diagnostic engine.
    3. A deterministic script processes the runs to calculate reliability metrics and cluster errors.
    4. You receive a formatted Markdown report detailing the findings, error clusters, and prioritized items for review.

    Frameworks & tools

    This skill is framework-agnostic and works with data from GitHub Actions, GitLab CI, Jenkins, CircleCI, or any system that exports step-level status and duration data.

    Why this beats prompting it yourself

    Standard LLM prompts often hallucinate performance metrics or miss subtle patterns in large log files. This skill uses a deterministic execution script to ensure failure rates and p90 durations are mathematically accurate while the agent handles the complex task of clustering unstructured error text.

    Use cases

    • Identify which tests are flaky vs. consistently failing.
    • Pinpoint why CI times have increased over the last month.
    • Group 100+ failed runs into 3-4 distinct root cause clusters.
    • Generate a reliability report for a team lead or engineering manager.

    Known limitations

    This tool is for historical diagnostic analysis only. It does not provide real-time monitoring, automatic pipeline fixes, or log parsing beyond structured error messages.

    Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for DevOps & Cloud.

    Does not perform automatic fixes or real-time monitoring. Limited to historical run data analysis. Requires structured step data (status, duration, error) to function accurately.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before buying a skill.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 1 day ago

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Environment Variables
    Listed1 day ago

    What's inside

    Frequently Asked Questions

    The skill requires structured pipeline data, such as JSON exports or logs containing step names, statuses, and durations. It works with data from GitHub Actions, GitLab CI, Jenkins, and CircleCI, provided the data captures individual step performance.

    Unlike standard prompts that might guess at performance, this skill uses a deterministic engine to calculate mathematically accurate failure rates and p90 durations. It combines this precision with LLM-powered error clustering to ensure you get reliable metrics alongside qualitative insights.

    This skill is designed for historical diagnostics and retrospective reporting rather than live monitoring. It is best used to analyze groups of recent runs to identify recurring patterns, flaky tests, or gradual performance regressions.

    You will receive the AI agent skill package, documentation on data formatting, and a structured report template. The purchase includes the logic required to process logs, cluster errors, and generate prioritized remediation recommendations.

    While this skill identifies the root cause and ranks issues by impact, it does not automatically modify your code or CI configuration. It provides a detailed Markdown report with specific recommendations so your team can implement the fixes safely.

    New in DevOps & Cloud