03-taskspawner-github-issues

03 — TaskSpawner for GitHub Issues

A TaskSpawner that polls GitHub for issues matching a label filter and automatically creates a Task for each one. This is the core pattern for autonomous issue resolution.

Use Case

Automatically assign an AI agent to every new bug-labeled issue. The agent clones the repo, investigates the issue, and opens a PR with a fix.

Resources

File	Kind	Purpose
credentials-secret.yaml	Secret	Claude OAuth token for the agent
github-token-secret.yaml	Secret	GitHub token for cloning, PR creation, and issue polling
workspace.yaml	Workspace	Git repository to clone into each Task
taskspawner.yaml	TaskSpawner	Watches GitHub issues and spawns Tasks

How It Works

TaskSpawner polls GitHub Issues (label: bug, state: open)
    │
    ├── new issue found → creates Task → agent fixes bug → opens PR
    ├── new issue found → creates Task → agent fixes bug → opens PR
    └── ...

Steps

1. Edit the secrets — replace placeholders in both secret files.

2. Edit workspace.yaml — set your repository URL and branch.

3. Apply the resources:

kubectl apply -f examples/03-taskspawner-github-issues/

4. Verify the spawner is running:

kubectl get taskspawners -w

5. Create a test issue with the bug label in your repository. The TaskSpawner picks it up on the next poll and creates a Task.

6. Watch spawned Tasks:

kubectl get tasks -w

7. Cleanup:

kubectl delete -f examples/03-taskspawner-github-issues/

Customization

• Change labels in taskspawner.yaml to match your labeling scheme.
• Add excludeLabels to skip issues that need human input (e.g., ["needs-triage", "wontfix"]).
• Adjust pollInterval inside the source block to control how often GitHub is polled.
• Set maxConcurrency to limit how many Tasks run in parallel.
• Edit promptTemplate to give the agent more specific instructions.