Move from alert to evidence-backed triage faster.
Trace a production error, identify the likely change, and prepare a triage issue without giving the agent authority to modify production.
How this composition works
This workflow brings error context, repository history, issue tracking, and edge signals into one isolated investigation. The agent develops and records a hypothesis, links every source, and prepares the next action. Engineers retain control over code changes, deployments, and incident declarations.
Read the failure
Collect the error, trace, environment, and affected release context.
Inspect the likely change
Compare the failure window with commits and pull requests without modifying the repository.
Check service boundaries
Rule in or out edge configuration and traffic anomalies.
Prepare the triage issue
Create a structured issue only after the evidence and uncertainty are explicit.
Connected systems
These are the real product connectors composed for this example, each scoped to the credential permissions your workspace approves.
Sentry
Investigate errors, traces, releases, and project health in Sentry.
Explore connectorGitHub
Read and write GitHub repos, issues, and pull requests.
Explore connectorLinear
Browse and update Linear issues, projects, and cycles.
Explore connectorCloudflare Developer Platform
Manage Cloudflare accounts and KV storage namespaces, and list Workers.
Explore connectorOptional skills
Where relevant, skills add reusable method and verification. The connectors and agent instructions remain the primary building blocks.
Try asking
More examples for Engineering
Build a governed workflow like incident triage.
Tell us what your team repeats today. We will map the systems, approval boundaries, and first governed run with you.