agentleFS
Microsoft AzureKnown publisherAzure/azure-sdk-tools/.github/skills/azsdk-common-pipeline-analysis/SKILL.md

azsdk-common-pipeline-analysis

Skill135 starsChanged 2 months agoMITView on GitHub

Analyze Azure SDK CI/CD pipeline failures into a structured diagnosis, and define the required output format. Load this skill before calling azsdk_analyze_pipeline, which returns raw failure data that this skill interprets and formats. USE FOR: "pipeline failed", "build failure", "CI check failing", "tests failing in CI", "analyze pipeline", "debug SDK pipeline". DO NOT USE FOR: local build issues without pipeline context, API design review, SDK publishing, applying code fixes (instead use azsdk-common-pipeline-fixer). INVOKES: azure-sdk-mcp:azsdk_analyze_pipeline, azure-sdk-mcp:azsdk_get_pipeline_llm_artifacts, azure-sdk-mcp:azsdk_get_failed_test_run_data, azure-sdk-mcp:azsdk_get_failed_test_case_data, azure-sdk-mcp:azsdk_get_pr_checks, azure-sdk-mcp:azsdk_get_pipeline_status.

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 0 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run azsdk_analyze_pipeline first, then categorize each failure and cite specific files/lines.

    line 20
  2. Very specific:

    Analyze: - Run azsdk_analyze_pipeline.

    line 37
  3. Very specific:

    Report: - Use the output format: root cause, affected files, per-failure fix + verify command, and fixable vs infrastructure.

    line 42
  4. Very specific:

    If azsdk_analyze_pipeline returns no data, verify the build ID is correct and the run has completed.

    line 52
  5. Very specific:

    If failed_pipeline_tests is empty, rely on failed_pipeline_tasks log analysis (test-artifact upload may not be configured).

    line 53

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Pipeline Analysis

This skill analyzes Azure SDK CI/CD pipeline failures and provides a structured diagnosis including root cause, affected files, and concrete instructions for how to fix each issue. It does NOT apply fixes - it tells you exactly what's wrong and how to resolve it.

Rules

  • Load this skill before running azsdk_analyze_pipeline — the tool returns raw failure data that this skill interprets and formats per output format.
  • Requires the azure-sdk-mcp server; without it, inspect logs in the Azure DevOps UI.
  • Analysis-only: never edit files or apply fixes — use azsdk-common-pipeline-fixer for changes.
  • Run azsdk_analyze_pipeline first, then categorize each failure and cite specific files/lines.
  • For infrastructure failures (network timeouts, agent crashes, throttling), recommend retry, not code changes.

MCP Tools

ToolPurpose
azure-sdk-mcp:azsdk_analyze_pipelineAnalyze pipeline failure (logs + test results)
azure-sdk-mcp:azsdk_get_pipeline_llm_artifactsDownload test result artifacts from pipeline
azure-sdk-mcp:azsdk_get_failed_test_run_dataFull details for every failure in one artifact file
azure-sdk-mcp:azsdk_get_failed_test_case_dataFull details for a single failing test in a file
azure-sdk-mcp:azsdk_get_pr_checksGet pipeline/check results linked to a PR
azure-sdk-mcp:azsdk_get_pipeline_statusGet pipeline run status

Steps

  1. Identify - Get the build ID, pipeline URL, or PR link (use the PR link when triggered from a PR comment).
  2. Analyze - Run azsdk_analyze_pipeline. It returns failed_pipeline_tasks (log errors from failed steps) and failed_pipeline_tests, a lightweight index of failed tests grouped by artifact file. Each entry has artifact_file_path, platform, and failed_test_titles — titles only, no error text, to keep the response small.
  3. Fetch failure details - Fetch failure details lazily so you don't overload context: call azsdk_get_failed_test_run_data with the parameter failedTestRunsPath set to artifact_file_path once per file to get every failure in it. Or call azsdk_get_failed_test_case_data with failedTestRunsPath set to artifact_file_path and testCaseTitle set to one exact title from failed_test_titles. Prefer the per-file call when triaging a whole file's failures.
  4. Categorize each failure: test, build/compilation, validation/lint, or infrastructure.
  5. Diagnose - Give each failure's root cause and affected file(s)/line(s), and note if several share one root cause. See failure patterns.
  6. Report - Use the output format: root cause, affected files, per-failure fix + verify command, and fixable vs infrastructure. Recommend azsdk-common-pipeline-fixer to apply fixes.

Examples

  • "My pipeline build 6447834 failed, what went wrong?"
  • PR comment: @copilot Analyze the failed pipeline on this PR
  • "Why is my CI red? Build ID is 6455939"

Troubleshooting

  • If azsdk_analyze_pipeline returns no data, verify the build ID is correct and the run has completed.
  • If failed_pipeline_tests is empty, rely on failed_pipeline_tasks log analysis (test-artifact upload may not be configured).
  • For a private/internal pipeline, the user may need to az login. See failure patterns.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.