Skip to main content

Smart Diff

Traffic Capture​

Smart Diff captures and analyzes HTTP traffic between services to detect breaking changes. It compares requests and responses between baseline and sandbox environments using AI models trained on historical traffic patterns.

We extend the starlet http module to accept two optional flags for each request: capture (type bool) and reqName (type string).

For example:

resp = http.get("...", capture=True, name="getLocations")

When capture is true, the library will inject headers based on environment variables to interface with our system of traffic capture. In this case, one can name the request with reqName.

Am example Smart Test is shown below:

res = http.get(
    url="https://my-api:8081/locations",  # supports internal URLs (http://name.namespace.svc)
    capture=True,
    name="getLocations"
)

print(res.status_code)
print(res.body())

When capture=True:

  1. Headers are injected for traffic capture
  2. Requests execute against both baseline and sandbox
  3. Responses are analyzed for structural differences
  4. Results are stored and accessible via dashboard

Analysis Output​

Smart Diff generates a relevancy-based comparison highlighting:

  • Schema changes
  • Field additions/removals
  • Type modifications
  • High-impact differences

Smart Diff Results

More details are available in the concept documentation.