Run QA agents to search your app for bugs
Give an agent your app and test rules, and let it probe flows, file reproducible bugs, and flag risky fixes for review.
- What they did
- Momentic’s Mo is a QA agent that runs in the cloud on web, Android, and iOS targets. Teams describe flows in plain-English YAML, add business rules to a Knowledge base, and connect MCP servers plus secrets from their org. From there, the explore agent can be launched from the CLI to navigate the product, write up bugs with repro steps, and feed failures into triage agents that suggest repairs for a person to check.
- How it went
- Momentic says its system has found 117,010 bugs. The vendor points to cases such as a 2FA step being bypassed and stored cards disappearing during checkout. It also says Quora reduced a daily test run from 7 hours to 30 minutes.
- Tips
- Keep the approval loop in place for any repair or action that changes data. Treat the generated diffs and triage suggestions as drafts until a person reviews them.
Try it yourself with Claude in Chrome
Test [app URL] with the [test account] and inspect the [user flow]. Act like a QA tester: move through the normal path, then try edge cases such as empty fields, bad input, back navigation, refreshes, and repeated taps or clicks. For every problem you find, give me the steps to reproduce, what should have happened, what actually happened, and a severity level. If the next action would charge money, remove data, submit something, or do anything irreversible, stop and ask for my approval before continuing.
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Check a codebase for redundant packages and polyfills
An agent can scan code and install commands to find dependencies that can be replaced with built-ins or better options.
Turn a docs site into Markdown files with an agent
They used a custom crawler so Claude could save a 60-page docs site as Markdown without crowding the chat.
Convert a spoken brief into a working task tree
A developer fed a spoken project brief into LLM agents, which turned it into a spec, plan, scaffold, module tickets, and PRs.
Build a terminal calling tool with human backup
They wired a CLI agent to place calls and loop in a person only when it needed an answer.
Assemble a R coding agent from three basic tools
They wired a Claude agent in R to file and shell helpers, then hardened it with safer path checks and editing functions.
Track agent memory in git for coding work
Store project context in git so a coding agent can reuse decisions and notes across sessions.