https://t.co/uD4EPqm2r7
The rchub-qa was built as part of ratl release intelligence efficiency for one of biggest retail chain and inspired by Context-Hub to opensource it.
[1] AI agents waste 50,000 tokens reading Playwright docs.
They need 2,000.
We just open-sourced rchub-qa — curated, agent-optimized QA docs that cut token usage by 94%.
22 docs. 15 tools. Zero fluff.
Playwright, Selenium, Cypress, Appium, k6, Locust, Pact, DeepEval & more
[3] Without rchub-qa
→ 500,000 tokens on doc retrieval
→ 40 repeated lookups
→ Same gotchas rediscovered every session
→ ~$10 in Claude Opus costs just for docs
With rchub
→ 10,000 tokens
→ 1 fetch, everything included
→ Gotchas saved forever via annotations
→ ~$0.20
Today, We are revealing how https://t.co/iUfniFbcKd actually works.
It’s not just another test tool. It’s a fully agentic, LangGraph-powered platform that reads your PRs, understands your architecture, generates & runs tests — autonomously.
@farooqadam@sreeramanmg@navi4ckm
@Austen The more reliable use case we’ve found is voice-to-testing workflows! We’ve run a few initial PoCs and are seeing significant success in areas like test execution, bug reporting, retesting, and sign-offs. It’s currently in the refinement phase for market fit