AeonDave schemathesis
Schema-driven API fuzzing for developer and AppSec workflows.
Quick Start
# CLI
uvx schemathesis run https://example.schemathesis.io/openapi.json
# Installed CLI
schemathesis run https://your-api/openapi.json
import schemathesis
schema = schemathesis.openapi.from_url("https://your-api/openapi.json")
@schema.parametrize()
def test_api(case):
case.call_and_validate()
Operator Flow
- First run with defaults to map baseline failure landscape.
- Triage in this order: undocumented status codes -> schema conformance -> server errors.
- Narrow scope by tag/path to fix incrementally.
- Run longer optimization profile for release/security gates.
- Keep CI smoke + nightly deep profiles separated.
Strengths
- Generates large input space from API schema automatically.
- Detects 500s, contract/schema drift, validation bypasses.
- Supports stateful API workflows (operation sequences).
Optimization Profile (Deep Runs)
schemathesis run <schema_url> \
--max-examples 1000 \
--continue-on-failure
Often paired with targeted generation and health-check suppression on complex schemas.
Practical Tricks
- Use include-path/include-tag scope to eliminate triage noise.
- Keep one strict schema-conformance pass and one exploratory bug-hunting pass.
- Export machine-readable artifacts (JUnit/Allure) for team triage.
- Re-run minimal reproducer from failing example before opening defect ticket.
Common Pitfalls
- Trying to fix every first-run failure globally at once.
- Ignoring repetitive undocumented status patterns that should be schema-level fixes.
- Treating long-run health-check warnings as always ignorable.
CI Pattern
- Run fast smoke profile on PRs.
- Run longer stateful/negative campaigns nightly.
- Publish JUnit/Allure artifacts for triage.