Chatoshi QA
QA Automation Journey

Chatoshi Mobile
Automated Testing

Every release is now checked automatically. 17 key user journeys run start to finish on their own, with no manual steps.

Muhammad Shawaiz, QA Automation Updated 2026-09-17 Full-suite green on Chatoshi 1.72.0, 2026-09-17
The numbers
Concrete wins from the foundation phase. See per-test results
11
Accessibility improvements shipped to both apps
17/17
Key user journeys passing, fully automated
10 → 0
Tests that broke when design shifted. They find buttons by name now
2
Apps covered, Chatoshi and VIP, from the same work
Shipped: 11 accessibility merge requests
Each MR added stable identifiers to a widget. Shared package means both apps benefit.
!1306
Message controls bar
Overflow menu on every AI response
!1308
Chat reason status widget
Report submission success sheet
!1309
Chat text field send button
Primary send or stop action in composer
!1315
Custom bottom nav bar
Chat, History, Alerts, Learn, Menu tabs
!1318
Scroll to bottom button
Chat surface auto scroll arrow
!1319
History item
Chat history list row
!1326
Rendering issue reminder
First time educational popup
!1327
First session explanation
Post onboarding starter suggestions
!1370
Onboarding skip button and chat input field
Priority 1 accessibility items
!1377
Onboarding next button
Priority 2 accessibility item
!1385
Response overflow onTap wiring
Unblocks Flow 13's id based tap
Roadmap
Four phases showing where we are and where we're going.
Done
Phase 1
Foundation
  • Full suite green, 17 flows automated
  • 11 MRs shipped
  • Manual login step eliminated
  • Coord taps eliminated
Now
Phase 2
Improve
  • Detect service outages early
  • Sort failures by cause
  • Team escalation playbook
  • Cleaner test selectors
Next
Phase 3
Expand
  • More coverage
  • New feature areas
  • Better selectors
  • Continued cleanup
Future
Phase 4
Scale
  • Broader reach
  • Deeper insights
  • Shared patterns
  • Nice to have polish
Six pillars of the framework
What holds the whole thing up.
How tests find things
By the button's name, not where it sits on screen. That is what stops tests breaking when design changes.
Coverage strategy
Smoke gates releases. Regression catches drift. Feature MRs ship with tests.
Reliability
Clean sim every run. Skip broken chains. Loud on real bugs.
CI/CD integration
Tests run on every MR. Failures block merges. Green means truly green.
Cross app scale
Shared package pattern where every win benefits Chatoshi and VIP together.
Observability
Structured logs. Failure reports. Trends visible to the whole team.
Auto trigger modals, status
Designer flagged 5 modals for automation. Current progress:
1
Rate app request
Native OS modal, outside Flutter's reach
Out of scope
2
Rendering issue reminder
First time popup after chat response
Shipped
3
First session explanation
Post onboarding starter suggestions
Shipped
4
Update available
Full screen new version prompt
Pending dev
5
Critical update available
Forced update variant
Pending dev
How the team makes it work
Automation scales when everyone owns their piece.
QA (me)
Owns flows, selector strategy, runner infra, and coverage roadmap.
Frontend devs
Add Semantics ids to new widgets as part of Definition of Done.
Backend devs
Provide testable states or debug flags where features can't otherwise be triggered.
Reviewers
Confirm identifier naming and public const exposure on every MR.
Team lead
Owns CI resources, merge policies, and cross team alignment.
Designer
Flags auto trigger UX to QA proactively so tests plan around it.

The message in one line

We've built a stable foundation for mobile E2E automation. From here we keep improving it, step by step, based on what the team needs most.

Chatoshi Mobile QA Automation, Framework Dashboard v1.0