Abstract: To reduce soil compaction in wheat fields on the North China Plain, and to address the problems that most existing small mobile platforms are developed for a single operation and that wheeled chassis carrying variable operation units are prone to instability on slopes and uneven ground, a small general-purpose autonomous mobile platform was developed for fixed-track seeding and subsequent field management operations. According to the agronomic requirements of wheat, the platform adopted a lightweight frame, four-wheel drive, four-wheel steering, a stepless lifting mechanism, and a quick-hitch mounting frame. The whole machine consisted mainly of a frame, drive hub motors, steering servo motors, electric lift actuators, a mounting frame, a navigation system, a control system, and replaceable operation units. The overall dimensions were 2070 mm × 1540 mm × (1170–1620) mm, and the ground clearance could be adjusted from 370 to 820 mm according to crop height. Different operation units could be mounted by replacing the whole mounting frame or by adjusting individual components connected with U-bolts, which improved the adaptability of the platform to seeding, soil-loosening, and other field management operations. A modular control system with dual Controller Area Network channels was constructed. One channel coordinated chassis motion and lifting, while the other was reserved for the mounted operation units, enabling better integration of operation units and coordination with vehicle motion. Ansys simulation was used to check the strength and deformation of the frame under two limiting working conditions, and the results showed that the frame met the structural reliability requirements for field operation. To solve the instability problem caused by changes in the center of gravity after different operation units were mounted, a theoretical model of the segmented rollover process on a three-dimensional slope was established. The model described the transition from four-wheel support to three-wheel support, clarified the wheel load transfer law, and proposed warning threshold and rollover threshold criteria under coupled pitch and roll conditions. A multibody dynamic model was further established in RecurDyn for verification. The simulation results agreed well with the theoretical predictions, with an average angular deviation of -0.7° and a root mean square error of 0.88°, indicating that the model could be used to predict the support state of the platform during field operation. For autonomous operation, a Π-shaped fixed-track operation path was planned using Real-Time Kinematic Global Navigation Satellite System (RTK-GNSS) positioning. The Stanley path-tracking algorithm was improved by introducing a lateral error smoothing term and a minimum speed constraint to improve the tracking accuracy during starting and non-uniform-speed operation. The optimized parameters were determined as a lateral gain of k=1.8, a weight coefficient of η=0.75, and an operating speed of v=0.8 m/s. Vehicle performance tests were conducted on cement pavement and soil ground, and seeding and field management function tests were carried out in wheat plots. The straight-line lateral offset rates of the platform on cement pavement and soil ground were 1.52% and 1.35%, respectively, and the inner turning radii were 0.69 and 0.99 m, respectively. The platform could climb a 16° slope and pass over a 10 cm obstacle, showing good trafficability and maneuverability in small plots. During autonomous navigation, the average lateral tracking error during the whole seeding operation was 0.049 m, and the average lateral tracking error in the uniform-speed operation stage was 0.026 m. The seeding test showed that the platform operated stably, with a maximum coefficient of variation of sowing depth of 10.10% and a coefficient of variation of seed discharge of 1.61%. The field soil-loosening test was conducted by reusing the seeding trajectory. The seedling injury rate was less than 2%, and the seedling crushing rate was less than 3%, indicating that the platform had good trajectory reuse performance for field management after seeding. The developed platform showed good adaptability to implement mounting, slope stability, path-tracking accuracy, and potential for modular expansion. This study can provide a reference for the design of lightweight and general-purpose intelligent platforms for wheat seeding and field management operations.