The principle of a self intelligent internal bio-micro nanopid, can be started by a minuet micro-mainframe.
The is a sub cell computer, that has drains and laminations performed, at atomic or near atomic sizes.
*The principle of ant self intelligence takes over, once these bots are activated.
The external shell is important next, as the shell, is a backwards affectation, which houses cellular miotic matter, much the same as an insect does.
There are a set of servo motors, which impart from the central brain mainframe, to cilia, on the outside of the micro-nanobots, which instructs, those nanobots, to travel to an appointed area, within the body.
Each bot, is taught to para to a human cell location, so that it is possible to establish symbeology, with the cellular grouping, then perform either construction or rearranging of internal body structures.
A plethora of nonreactive mentals and plastics, could be used to construct each bot.
A recall signal, could be issued in code, to either stop the progress of the nanobots, recall them to excretion, or temporarily deactivate them, for a period of time.
Hive intelligence is the key intent here, as this is a collection of collective miniature minds, rather than one sole bot.