Project

sdet_flow

0.0
The project is in a healthy, maintained state
sdet_flowsdetflow/sdetflowHomepageDocumentationSource CodeBug TrackerWiki
Ruby automation utilities for API testing, retries, redaction, web adapter fallbacks, and integration with GenAI-powered quality engineering tools.
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
 Popularity
Downloads
1,064
Stars
0
Forks
0
Watchers
0
 Releases
Current version
0.2.0
Total releases
1
First release
2026-09-28
Latest release
2026-09-28
 Issues
Open Issues
2
Closed Issues
3
Total Issues
5
Issue Closure Rate
60%
 Pull Requests
Open Pull Requests
3
Closed Pull Requests
8
Merged Pull Requests
0
Pull Request Acceptance Rate
0%
 Development
Primary Language
TypeScript
Licenses
Apache-2.0
Average date of last 50 commits
2026-09-29
Reverse Dependencies
0

 Dependencies
 Project Readme

SDETFlow

SDETFlow CI Playwright Browser Matrix Public npm Consumer Smoke Public RubyGems Consumer Smoke Public Maven Consumer Smoke

Open-source test automation, GenAI-powered quality engineering, and Agentic AI tools by Sumanth Gumedelli.

Current release: SDETFlow v0.2.0

SDETFlow is a multi-language quality-engineering ecosystem focused on reusable automation architecture rather than one-off test scripts. The v0.2 line contains independently consumable TypeScript/Playwright, GenAI + Agentic AI, test-intelligence, Java API, and Ruby automation libraries with deterministic behavior, privacy controls, packaging checks, and CI-ready documentation.

Release status: all five SDETFlow v0.2.0 libraries are publicly published: three npm packages, the sdet_flow RubyGem, and the Java sdetflow-api artifact on Maven Central. Clean public-registry consumer install/import/compile smoke tests are green, along with GitHub CI, the Chromium/Firefox/WebKit browser matrix, packaging checks, and live OpenAI/Gemini provider contracts.

Packages

Package Purpose Version
@sdetflow/playwright Playwright diagnostics, smart locator fallback, evidence capture, retries, test data, optional AI hooks 0.2.0
@sdetflow/ai Provider-agnostic GenAI + Agentic AI quality engineering SDK with OpenAI/Gemini adapters, routing, budgets, analyzers, registered tools and approval gates 0.2.0
@sdetflow/insights Playwright/JUnit/Cucumber/Allure normalization, flake scoring, clustering, trends and portable persistence 0.2.0
io.github.sumantthh-sdet:sdetflow-api Java 17 API automation SDK with safe retries, auth, JSON assertions and telemetry 0.2.0
sdet_flow Ruby quality-engineering Gem with API automation, retries, redaction and Capybara/Watir adapters 0.2.0

Engineering principles

  1. Deterministic first. Core tests do not depend on an LLM being available.
  2. AI is optional and advisory. GenAI can analyze evidence, but does not silently change test outcomes.
  3. Agentic side effects are host-controlled. Models may propose registered tools; the application owns validation, authorization, execution and approval.
  4. Secure defaults. Common credentials/tokens are redacted, sensitive artifacts are bounded or opt-in, and provider keys stay outside prompts.
  5. Safe retries. Unsafe operations are not retried by default.
  6. Explainability. Locator fallback, retries, failure clusters, AI routing and agent steps emit evidence/telemetry.
  7. Consumer usability. Release verification installs generated artifacts into clean temporary consumers before a release is considered ready.

Agentic Quality Engineering

@sdetflow/ai includes a bounded, host-controlled Agentic AI runtime for quality-engineering workflows. It supports registered-tool allowlists, strict JSON decisions, tool input validation, execution timeouts, maximum-step limits, untrusted-output boundaries, secret/PII redaction, and explicit human approval for high-sensitivity tools. It does not grant models arbitrary shell access or bypass application authorization.

Typical use cases include evidence-driven failure investigation, test-result analysis, safe workflow orchestration, and human-governed quality actions. MCP/CI/Jira adapters can be layered on the public tool contract without changing the core orchestration model.

Local verification

./scripts/verify-all.sh
./scripts/verify-release.sh

The first command runs package test suites and website checks. The second builds/packs installable artifacts and validates them from clean consumer projects.

Documentation

Start with:

  • docs/ARCHITECTURE.md
  • docs/AGENTIC-AI.md
  • docs/BUILD-STATUS.md
  • docs/test-specs/MASTER-TEST-SPEC.md
  • docs/PUBLISHING.md
  • SECURITY.md

Author

Sumanth Gumedelli — Senior SDET / QA Automation Architect building scalable test platforms, open-source automation libraries, and GenAI-powered quality engineering tools with Agentic AI workflows.

License

Apache License 2.0.