This directory publishes reusable GitHub Agentic Workflows
that can be installed into other repositories. The workflow sources are kept outside
.github/workflows/ so they do not run in this repository.
| Package | Description |
|---|---|
| Build Failure Analysis | Analyzes failed .NET Azure Pipelines builds from their existing binary logs and posts evidence-backed findings on the pull request. |
| MSBuild Quality Review | Reviews changed MSBuild project and infrastructure files and posts one bounded, evidence-backed PR review comment. |
| Test Failure Analysis | Analyzes bounded, normalized test evidence from a repository-owned CI collector and reports supported failures, flakes, hangs, crashes, and duration regressions. |
| Unskip Closed Tests | Re-enables source-bound .NET tests only after deterministic tracking-state, revision, edit, and structured execution verification. |
Install the complete collection:
gh aw add-wizard dotnet/skills/agentic-workflows@mainOr install an individual package using the command in its README. Pin production
installations to a release tag or commit SHA instead of main.
Installed workflows are tracked by source: metadata. Pull compatible updates into a
consumer repository with:
gh aw updateThe installer copies each workflow source and its dependencies into the consumer
repository, then generates the executable .lock.yml file there. Generated lock files
are therefore not stored in this distribution directory.
Every package has a Vally-format spec at
tests/agentic-workflows/<package>/eval.yaml. The normal /evaluate discovery
includes package sources, resources, shared graders, and these scenarios.
Scheduled evaluations include the collection; manual evaluation dispatch with
plugin: agentic-workflows selects only these packages.
For a local run:
dotnet run --project eng/skill-validator/src/SkillValidator.csproj -- evaluate `
agentic-workflows/msbuild-quality-review/aw.yml `
--tests-dir tests/agentic-workflows --runs 1 --verdict-warn-onlyThe native SDK lane loads real local imports and installed package resources,
compares them with a no-workflow baseline, and retains a package-agent arm.
Results are published through the normal pipeline with skillKind: workflow
and evaluationLane: workflow-prompt-sdk.
These are offline prompt/decision evaluations: fixture evidence replaces
collectors and external services, and result.json contains proposed actions.
They do not execute Actions bootstrap jobs or publish safe outputs. Compilation,
trusted-helper tests, runtime trace graders, and consumer-repository integration
runs cover different contracts and remain separate evidence.