Compare automation approaches
Put input replay and state-aware modular automation side by side through practical tasks.
This is not a review of specific competing products. “Recording and playback” describes that method alone; it does not claim every other macro tool is limited to it.
| Task / need | Recording and playback alone | JAQ Macro approach |
|---|---|---|
| Repetition | Replays recorded inputs in order. | Adds observation, conditions and result verification to recorded actions. |
| Target position changes | A fixed point may need to be recorded again. | Use located targets from image search or object tracking. |
| Object appearance changes | Input coordinates alone do not recognize an object. | VOT exemplars, YOLO11 ONNX and hybrid options, with suitable data and testing. |
| Prepare training data | Outside the scope of input recording. | Annotation, GrabCut/Alpha, 360° Copy-Paste, shadows and appearance options. |
| Use visible text | Needs an additional way to read and interpret text. | Local Windows/Paddle OCR, optional AI OCR and variable transfer. |
| Reach world coordinates | The same input duration does not verify the same destination. | Coordinate readings, heading, axes, tolerance and waypoints provide feedback. |
| Decisions and loops | Follows the recorded sequence. | True/false branches, nested conditions, bounded loops and 32 functions. |
| Large-project structure | A long recording can become difficult to maintain. | Groups, subgroups, calls, returns, transfers and shared checks. |
| Unexpected states | An unrecorded state is not evaluated by input replay alone. | Window/image supervisors, recovery groups and practical limits. |
| AI-assisted design | Input recording alone does not expose a module contract. | MCP project tools, module information, a central guide and design tasks. |
| Backup and sharing | Depends on the tool and recording format. | Relevant Pro rights enable group selection, dependencies and linked PNGs. |
| Testing and diagnosis | Observe what happens during playback. | Dry run, step/group tests, logs, result variables and real-state verification. |
Compare on a real task
Use the same target and scenario. Evaluate preparation time, false actions, reaction to changed state and maintenance effort together. A listed feature does not prove correct behavior with every configuration.
What does JAQ Macro focus on?
Tasks combining visual observation, conditional workflows, coordinate feedback and MCP-assisted design. Begin with a suitable small module instead of adding unnecessary complexity.