Two AI Agents That Bookend QA on monday.com | LupApps Pod 02

I built two AI agents to bookend my QA tester. They do everything that can be automated. The human tester does only what proves it works.

🔍 Lara Croft (Test Plan Generator) — when a feature flips to "Ready for QA," she reads the spec, maps the unknown terrain (edge cases, regressions, race conditions), and generates up to 25 test cases as subitems on the QA board.

🛡 Gatekeeper (Release Reviewer) — when the human tester finishes, Gatekeeper scores GO / CONDITIONAL / NO-GO against fixed thresholds. Any Critical bug = automatic NO-GO. No exceptions.

In this video you'll see:
0:00 — A feature flips to "Ready for QA"
0:20 — Lara Croft generates ~20 test cases as subitems + pings Slack
0:50 — The human tester fills results: 18 Pass, 1 Critical Fail
1:10 — Gatekeeper wakes up, scores the release as NO-GO
1:40 — Slack verdict cites specific test IDs + the Critical bug
2:00 — Why this works: the bots own everything mechanical, the human owns judgment

⚙️ Architecture
• Brain + 2 Skills on monday.com Agent Builder
• Boards: Team Tasks (trigger) + QA Plans (writes) + Slack (output)
• Knowledge: qa-standards.md (test format, severity guide, component checklists, pass thresholds)
• Triggers: status change → "Ready for QA" + completion update
• Failure isolation by design — each skill is self-contained

💰 Business impact (LupApps team — 4 people, 3 live monday apps)
~25 tests per feature · ~$35K/year recovered in time + prevented Critical-in-prod incidents

Built by LupApps. monday.com Partner. monday-agents.lupapps.io
Live deck of all our agents: https://monday-agents.lupapps.io

Want LupApps to build agents for your monday.com team?
→ idan@lupapps.io

#mondayagents #mondayAgentBuilder #AIagents #LupApps #QA #softwarequality #automation #aitools

Видео Two AI Agents That Bookend QA on monday.com | LupApps Pod 02 канала LupApps