Agentic testing
Mobot in the Agentic Software Factory
How coding agents, Mobot AI and Mobot testing robots on real devices fit together in the Agentic Testing Protocol, from test request to verified fix.
Client
Agents and tools
- Coding agents
- CI/CD pipeline
- Jira / Linear
Agent interface
Agent gateway
- MCP Server (read-only)
- Test Intent Schema
- Agentic Testing Protocol
Test intelligence
Mobot AI
- Test Planning and Coverage Selection
- Test Case Generation
- Tests that update as your app changes
- Mobot AI Model
Robot fleet
Real devices, real robots
- On Demand Robot Reservations
- Automated Device Reset and Provisioning
- Device Configuration Profiles
- Mobot testing robots
Results to client agents
- Human QA validation
- Structured defect reports
- Jira / Linear ticket sync
- Agent verification re-test
Logs and exhaustive defect forensics
An architecture model, not a feature list. Available to agents today: read-only access through Mobot’s MCP server to test reports, test cases, observations, and device and network logs where captured.
The protocol
The Agentic Testing Protocol lifecycle
The Agentic Testing Protocol describes how an agent hands a test run to Mobot and gets results back, stage by stage.
Stage 1: Discover
In the protocol, the agent sees which real devices and OS versions are available.
Stage 2: Reserve
In the protocol, the agent books device time or joins the queue.
Stage 3: Provision
In the protocol, the build is handed off, and the device is wiped, configured and the app installed.
Stage 4: Direct
In the protocol, the agent states the objectives, scope and acceptance criteria.
Stage 5: Execute
In the protocol, a Mobot testing robot and an operator run the test while the agent follows progress.
Stage 6: Report
In the protocol, results and structured defect reports come back to the agent.
Stage 7: Verify
In the protocol, Mobot re-tests the specific defect on the new build.
New build? The loop starts again at Discover or Direct.
What agents can do today
Through Mobot’s MCP server, agents get read-only access to test reports, test cases, observations (defects and issues), and device and network logs where captured.
Device booking, provisioning, directing runs and re-tests are what the Agentic Testing Protocol is built toward.
Architecture
Layer by layer
1 · Client
Agents and tools
Coding agents, your CI/CD pipeline and Jira / Linear are the clients. In the protocol, agents send test work to Mobot, read results back, and tickets sync with Jira / Linear.
- Coding agents
- CI/CD pipeline
- Jira / Linear
2 · Agent interface
Agent gateway
Agents connect to Mobot through the agent gateway. In the protocol, the MCP server exposes reports, test cases, defects and logs, a Test Intent Schema describes what to test, and the Agentic Testing Protocol hands it to Mobot.
- MCP Server (read-only)
- Test Intent Schema
- Agentic Testing Protocol
3 · Test intelligence
Mobot AI
Mobot AI covers test planning and coverage selection, test case generation, and tests that update as your app changes, on the Mobot AI Model.
- Test Planning and Coverage Selection
- Test Case Generation
- Tests that update as your app changes
- Mobot AI Model
4 · Robot fleet
Real devices, real robots
Mobot testing robots run tests on real devices. In the protocol, agents reserve robot time on demand, with automated device reset and provisioning and device configuration profiles.
- On Demand Robot Reservations
- Automated Device Reset and Provisioning
- Device Configuration Profiles
- Mobot testing robots
5 · Results to client agents
Results back to your agents
In the protocol, test reports, structured defect reports and logs (where captured), human QA validation status, Jira / Linear ticket sync and agent verification re-tests flow back to agents.
- Human QA validation
- Structured defect reports
- Jira / Linear ticket sync
- Agent verification re-test
- Logs and exhaustive defect forensics
See it on your app
Get a real defect report from Mobot testing robots and QA analysts, on your app, on real devices.