Agent Qa

This product is Vostride's self improving QA agent for web, Android and iOS applications. It executes tests written in plain English against the live interface, resolving each action and assertion on screen rather than through hardcoded selectors, and keeps the test definitions as reviewable YAML inside the repository. It is built for engineering teams that already use coding agents and want those agents to verify their own changes. Tests can be authored by product managers, engineers, QA staff or a coding agent, then run through a CLI or an MCP server, with packaged skills for authoring, debugging and triaging results. Each run leaves evidence attached to its steps: the elements that resolved, the flow that worked and real timings. Successful runs become validated memory and cached action plans that make later runs faster, and a curator component learns only from valid evidence. Web tests run on Playwright and mobile tests on Appium, with suites, hooks, saved auth state and a local dashboard. It works with any hosted or local model and with Codex and Claude Code subscriptions. Vostride publishes no pricing page. The tool installs as a development dependency from npm, its source is on GitHub, and model usage is billed by whichever LLM provider you configure. Tests, configuration, memory and artifacts stay under the user's control in the repository.