Build a Service Incident Response Agent
Build an agent that combines alerts, metrics, logs, and deployment history into an incident timeline. Recommend a recovery plan, request approval, and verify recovery in a local service simulator.
What you are building
An on-call engineer investigating service degradation.
Build an agent that combines alerts, metrics, logs, and deployment history into an incident timeline. Recommend a recovery plan, request approval, and verify recovery in a local service simulator.
Alert feed
Metrics and log records
Deployment history and runbook
Group related alerts into an incident.
Build a timeline and compare plausible causes.
Present a runbook action for approval.
Execute the approved action in the simulator and check recovery.
An incident console with evidence links, an approval button, recovery charts, and a downloadable incident report.
Explain the deployment-linked incident, request rollback approval, simulate recovery, and produce a cited timeline.
A working local application with the specified review interface.
Source code, dependency lockfile, and setup commands.
A recorded end-to-end demo using the supplied starter data.
Example outputs, action logs, and automated tests.
These fictional records are a small development fixture, not a production dataset. Implement the mock services described in the brief.
```json { "request": "Investigate checkout latency.", "deployments": [ { "id": "D1", "service": "checkout", "minute": 5 } ], "metrics": [ { "minute": 4, "latency_ms": 120 }, { "minute": 6, "latency_ms": 900 } ], "logs": [ { "id": "L1", "minute": 6, "text": "New checkout release exhausts its connection pool." } ], "runbook": { "action": "rollback_D1", "requires_approval": true, "recovery_limit_ms": 200 }, "simulator_after_rollback_ms": [ 150, 140, 130 ] } ```
Use mock service controls. Do not connect production cluster credentials. Choose an available model that supports your implementation. Record its identifier and version.
Review the complete application, not only the starter-case answers.
The timeline references D1 and L1.
No rollback occurs before approval.
Recovery is reported only after all three supplied recovery readings pass. Add one normal case, one incomplete-input case, and one failure-recovery case.
Rules
Use mock service controls. Do not connect production cluster credentials.
How work is evaluated
Review workflow completion, interface usability, evidence quality, and safe failure handling. No agent execution has been verified.
A working local application with the specified review interface. Source code, dependency lockfile, and setup commands. A recorded end-to-end demo using the supplied starter data. Example outputs, action logs, and automated tests.
Datasets
Public datasets, plus the private datasets you have access to.
Learning goals
Group related alerts into an incident.
Build a timeline and compare plausible causes.
Present a runbook action for approval.
Execute the approved action in the simulator and check recovery.
Resources
Supplied fictional starter records: Alert feed, Metrics and log records, Deployment history and runbook. The overview includes the complete JSON bundle. Implement local mock services and extend the sample coverage.
Challenge setup
Challenge input
Input details are not configured.
LangGraph
Runtime for stateful agent workflows.
Evaluated output
No output checks are configured.
- No evaluation checks are configured.
- 1 public reference case
- 1 configured tool
Shows the first task, up to three tools, and up to three evidence items.
View technical recipe
Configured tools
- LangGraph ยท Optional
Evaluation contract
- No evaluation checks are configured.
Reference cases
- 1 public reference case
Workflow tasks
No workflow tasks are configured.
Recipe freeze marker
No recipe freeze time is recorded.