Find agent skills by outcome

- Include visual trend charts showing latency

throughput

- Run chaos tests in staging first

then graduate to production with limited blast radius

- Test edge cases: empty responses

maximum payload sizes

- [ ] Security testing completed: injection

authentication

- Use realistic user behavior patterns

not synthetic uniform requests

- [ ] Resilience tested: circuit breakers

graceful degradation

- [ ] All endpoints tested under baseline

peak

- [ ] Response time percentiles (p50

p95

- Compile test results into structured report with metrics

bottlenecks

- Test webhook delivery

retry behavior

- Define success criteria: target response times

throughput

- Information disclosure in error responses: stack traces

internal paths

- Authentication weaknesses: token leakage

missing expiration

- Resource Utilization: CPU <70% at expected load

memory stable without unbounded growth

- Inadequate retry logic: no retries

or retries without backoff causing retry storms

- Error Rates: 5xx errors <0.1%

4xx errors <5% (excluding 401/403)

- Response Time: Simple GET <100ms (p95)

complex query <500ms (p95)

- Test proper error propagation: errors are meaningful

not swallowed or leaked as 500s

- Configure meaningful alerts based on SLI/SLO thresholds (e.g.

p95 latency > 500ms

- Identify resource bottlenecks (CPU

memory

- Profile endpoint response times at baseline load

capturing p50

Creative

- Design load test scenarios: gradual ramp

spike test (10x sudden increase)

PromptBeginner5 minmarkdownQuality: 28

- Execute load and stress tests by simulating realistic user behavior

gradually increasing load to find breaking points

PromptBeginner5 minmarkdownQuality: 28

- Profile endpoint performance by measuring response times under various loads

identifying N+1 queries