CentralRepoOps | GitHub Agentic Workflows
Skip to main content CentralRepoOps CentralRepoOps uses a central automation repository to host automation targeting software in one or more other repositories, or to aggregate their activity. It is a sub-pattern of MultiRepoOps . The repository can run workers, with or without an orchestrator, or collect events from component repositories. Using a Central Automation Repository Section titled “Using a Central Automation Repository” A central automation repository holds the workflows, schedules, and credentials; the software being improved stays in the target repositories.
Workers run in the automation repository and open issues or pull requests against those targets. An orchestrator is optional. flowchart LR subgraph central["Central automation repository"] tests[Worker Workflow A] perf[Worker Workflow B] orch[Optional orchestrator] -. dispatch .-> tests orch -. dispatch .-> perf end subgraph targets["Target software repositories (1..N)"] r1[Repository 1] r2[Repository 2] rn[Repository N] end tests --> r1 tests --> rn perf --> r2 perf --> rn Self-Scheduling Workers Section titled “Self-Scheduling Workers” The simplest deployment is a set of independent workers, each with its own schedule and target repositories.
For example, a test improver periodically adds missing tests in one repository, while a performance improver benchmarks and proposes optimizations in one or more repositories. Neither needs an orchestrator or dispatch workflow. Each worker defines its own on.schedule , reads or checks out the target software, and uses cross-repository safe outputs to propose changes. The workflows remain in the central repository; they do not need to be installed in the software repositories.
Optional Orchestration Section titled “Optional Orchestration” Add an orchestrator when work needs shared prioritization, target selection, or phased rollout across repositories. It dispatches workers in the central repository with a target repository as input. Workers can be dispatch-only or also retain their own schedules. The following example shows the orchestrated variant. It supports pilot waves and a central decision trail without installing workflows in each target repository.
Orchestrator ( dispatch-workflow safe output + max limit): --- on : schedule : weekly on monday tools : github : github-token : ${{ secrets.GH_AW_READ_ORG_TOKEN }} toolsets : [ repos ] safe-outputs : dispatch-workflow : workflows : [ worker-workflow ] max : 5 --- # Rollout Orchestrator Filter repositories, categorize by complexity, prioritize the rollout order, and dispatch the worker workflow for each selected repository.
Summarize candidates, breakdown, and rationale. Worker ( checkout + target-repo safe outputs per dispatched repo): --- on : workflow_dispatch : inputs : target_repo : description : ' Target repository (owner/repo format) ' required : true type : string checkout : repository : ${{ github.event.inputs.target_repo }} github-token : ${{ secrets.ORG_REPO_CHECKOUT_TOKEN }} current : true safe-outputs : github-token : ${{ secrets.GH_AW_CROSS_REPO_PAT }} create-pull-request : target-repo : ${{ github.event.inputs.target_repo }} max : 1 --- # Worker: Apply Changes to Target Repository Analyze ${{ github.event.inputs.target_repo }}, apply the required changes, and create a pull request explaining what was changed and why.
When using an orchestrator, keep its permissions narrow and delegate repo-specific writes to workers. Add correlation IDs to dispatch inputs for tracking. See the Dependabot Rollout example for a complete end-to-end walkthrough. Aggregating Events Section titled “Aggregating Events” Each component repository runs its own workflow that forwards events to the central automation repository via target-repo . The central repository accumulates a unified view without needing direct access to individual component repos. flowchart LR compA[Component repo A] -->|create-issue| central[Central automation repository] compB[Component repo B] -->|create-issue| central compC[Component repo C] -->|create-issue| central Useful for component-based architectures where multiple teams need a shared visibility layer, cross-project initiatives, or aggregating metrics from distributed repositories.
See Cross-Repo Issue Tracking for a complete example. Learn More Section titled “Learn More” MultiRepoOps — Side repo and downstream sync patterns Dependabot Rollout — End-to-end org-wide rollout example Cross-Repo Issue Tracking — Aggregated issue tracking example Cross-Repository Safe Outputs — Configuration reference Authentication — PAT and GitHub App setup