Where Testver is headed next
A clear view of what we’re building. Mobile automation is already in your hands in beta; autonomous AI agents and a direct line from your AI coding tools are next on the rail.
- Phase 1Beta · available now
Mobile Automation
Native Android & iOS, the same Testver workflow
Native Android and iOS testing through Appium — recorded, generated, run, and reported without ever leaving Testver. Tap through your app once and AI writes an idiomatic, self-healing test in your framework’s style, then runs it on real devices.
- First-class Appium adapters for Android & iOS
- AI Record for mobile with self-healing locators
- Cloud device clouds (BrowserStack, LambdaTest, Sauce Labs) and locally-connected devices
- Web / Mobile / Both workspace modes that adapt the UI to what you test
- Phase 2In progress
Autonomous AI Agents & Workflows
Give an agent a goal, not a prompt — and let it finish the job
Move beyond one-shot AI actions to agents that own an outcome end to end. Hand one a requirement or a failing suite and it plans its own steps, picks its own tools, and keeps going until the goal is met — generate the tests, run them, read what broke, fix the real cause, re-run to confirm. Agents judge their own results and self-correct between steps, run unattended on a schedule or a pipeline trigger, and stop to ask when a decision is genuinely yours. Every tool call stays permission-gated and auditable, so autonomous never means unsupervised.
- Goal-driven agents that plan their own steps and choose their own tools
- Self-correcting loops — run, cluster the failures, fix the root cause, re-run to confirm
- Reusable Agent Skills composed into saved multi-step Workflows your team can share
- Unattended execution on a schedule or CI trigger, with human checkpoints and a full audit trail
- Phase 3Planned
AI Coding Tool Integration
Drive Testver from Claude Code, Codex, Antigravity & Cursor
The AI tool you write code in should be able to test that code too. Testver will expose its own MCP server, so Claude Code, Codex, Antigravity, Cursor, and any other MCP-aware assistant can generate tests, run them, and read the results without you ever leaving the editor or terminal. Ask your coding agent to cover the feature it just built and it drives Testver end to end — every run still recorded in Testver, every action permission-gated, everything executing locally on your machine.
- Testver as an MCP server — one-step connect for Claude Code, Codex, Antigravity & Cursor
- Agent tools to scaffold, generate, run, and re-run suites in the framework you already use
- Failures returned as structured context — stack trace, diff, and trace links the agent can act on
- Generated project rules (CLAUDE.md, AGENTS.md, .cursor/rules) so agents follow your test conventions
Keep your framework. Add AI and a premium toolset.
Set up in minutes. Self-host on your own hardware. Describe a test in plain English and get working code in your own framework — in a fraction of the time.