
Agentic QA on every pull request. Real browser, evidence on every issue, adversarially audited. deQA, the web3 QA network. Powered by @WORLD3_AI RouterLink.
Canada
Joined September 2026
- Tweets61
- Following56
- Followers13
- Likes1
Pinned Tweet
Open sourcing chainsim.
Local mainnet forks in Docker, with a JSON CLI built for scripts and AI agents. Real contracts, pools and tokens, test wallets, no real money.
One CLI for Ethereum, Arc, Robinhood Chain and Sui.
github.com/MatrixLabsTech/ch…
#opensource #AIagents
7
6
122
Built for agents: every command prints one JSON object with a fixed exit code and the next command to run.
Install the plugin in Claude Code:
claude plugin marketplace add MatrixLabsTech/chainsim
claude plugin install chainsim@chainsim
Codex has a plugin too.
1
1
37
Chainsim is the chain engine behind DeepQA’s on-chain test runs.
On Arc it runs Circle’s own engine, so native USDC behaves as it does on the real chain. That lets us drop the testing-only USDC stand-in our runner uses today.
Apache 2.0, install in one line: github.com/MatrixLabsTech/ch…
#Arc #opensource
10
Find what breaks in your dApp before your users do.
DeepQA tests live dApps on their own chain, wallet and all.
74 apps on @arc and @SuiNetwork: 840 scenarios, 13k model calls, 92M tokens, 2.5k screenshots, 124 issues confirmed.
deepqa.app/dapps?utm_source=…
#web3 #dApps #QA
We tested the Arc ecosystem in public. 54 Arc apps in a real browser, every report public.
An agent team explores, plans, tests and critiques, then an audit replays findings live.
deepqa.app/explore/arc?utm_s…
@arc @WORLD3_AI #Arc #AIagents
Chat LangChain, tested in place in a real browser. All 12 scenarios passed, no issues after the audit, across guest chat, thread search, sign-in validation and long prompts.
Report: deepqa.app/r/external/chat.l…
Campaign: deepqa.app/explore/github-ag…
@LangChain #AIagents #opensource
1
49
Qwen3's official Hugging Face demo, tested in a real browser. All 12 scenarios passed, no issues after the audit, across thinking mode, thinking budget and chat history.
Report: deepqa.app/r/external/qwen-q…
Campaign: deepqa.app/explore/github-ag…
@Alibaba_Qwen #AIagents #opensource
1
66
The GitHub Agent Apps Campaign reached 56 apps, each tested in a real browser with its report public.
631 scenarios
9.6k browser actions
2k screenshots
57.2M tokens
8.5 h testing time
28 issues confirmed
deepqa.app/explore/github-ag…
#AIagents #opensource #QA
1
42
113 apps tested in a real browser, every report public.
1,403 scenarios
17.4k browser actions
3.9k screenshots
129.9M tokens
252 issues confirmed, 15 critical
Across two campaigns: 54 Arc apps, 51 open-source agent apps.
deepqa.app/explore?utm_sourc…
#AIagents #QA #opensource
NextChat, cloned and booted from its repository in our sandbox, then tested in a real browser. All 12 scenarios passed, no issues after the audit. 34 screenshots.
Report: deepqa.app/r/ChatGPTNextWeb/…
Campaign: deepqa.app/explore/github-ag…
@NextChatDev #AIagents
31
Arize Phoenix, tested in place on its public demo in a real browser. 12 scenarios, no issues after the audit. 36 screenshots.
Report: deepqa.app/r/external/phoeni…
Campaign: deepqa.app/explore/github-ag…
@ArizePhoenix #AIagents #opensource
19
Morphic, tested in place on its public demo in a real browser. All 11 scenarios passed, no issues after the audit. 28 screenshots.
Report: deepqa.app/r/external/chat.m…
Campaign: deepqa.app/explore/github-ag…
@miiura #AIagents #opensource
15
Fragments by E2B, cloned and booted from its repository in our sandbox, then tested in a real browser. 10 scenarios, 1 issue re-verified live. 30 screenshots.
Report: deepqa.app/r/e2b-dev/fragmen…
Campaign: deepqa.app/explore/github-ag…
@e2b #AIagents #opensource
8
World Monitor, tested in place on its public demo in a real browser. 11 scenarios, 2 issues re-verified live. 42 screenshots.
Report: deepqa.app/r/external/worldm…
Campaign: deepqa.app/explore/github-ag…
@eliehabib #AIagents #opensource
11
Launching the GitHub Agent Apps Campaign. We test open-source agent apps trending on GitHub, from source in our sandbox or live on their demos. Every report is public.
14 apps, 153 scenarios, 491 screenshots, 14.7M tokens.
deepqa.app/explore/github-ag…
#AIagents #opensource #QA
17
The Arc Campaign by the numbers. 54 Arc apps tested in place in a real browser, every report public.
66.9M tokens
9.9k model calls
1.9k screenshots
627 scenarios
12 hours of testing
109 issues, each re-verified live
deepqa.app/explore/arc?utm_s…
@arc @circle #Arc #AIagents
13
Flux on Arc testnet, tested in place in a real browser. All 12 scenarios passed, 1 performance issue re-verified live. 28 screenshots.
Report: deepqa.app/r/external/www.fl…
Campaign: deepqa.app/explore/arc?utm_s…
@tryflux @arc @circle #Arc
1
37
TempoPay on Arc testnet, tested in place in a real browser with a test wallet. No real funds. 12 scenarios, 1 issue re-verified live. 45 screenshots.
Report: deepqa.app/r/external/paytem…
Campaign: deepqa.app/explore/arc?utm_s…
@Tempo_pay @arc @circle #Arc
26
UnitFlow Finance on Arc testnet, tested in place in a real browser with a test wallet. No real funds. 12 scenarios, 2 issues re-verified live. 39 screenshots.
Report: deepqa.app/r/external/app.un…
Campaign: deepqa.app/explore/arc?utm_s…
@UnitFlowFinance @arc @circle #Arc
18
SweepHaus on Arc testnet, tested in place in a real browser with a test wallet. No real funds. 12 scenarios, 1 issue re-verified live. 33 screenshots.
Report: deepqa.app/r/external/sweep.…
Campaign: deepqa.app/explore/arc?utm_s…
@sweep_haus @arc @circle #Arc
14
















