Ten experiments, each with its species, its evidence tag and its PubMed id, and the three overclaims that would sink the argument if they were made.

First, what it is not

Bioelectricity is not electricity. Not current, not a spark, not a shock. It is a pattern of ions held across tissue, a standing map of charge that every cell maintains at rest and burns energy to keep, read by the cells that live in it. Nerve spikes are the fast, famous special case. The slow, quiet version in every other tissue is the one that says what to build.

The ten results the argument stands on

E1. The left-right body axis is set by a pump. An H+/K+-ATPase sits asymmetrically in the four-cell frog embryo and decides which side the heart goes on; omeprazole randomises it. Frog. EST. PMID 12372302.

E2. A second pump is required for the same axis. Blocking V-ATPase at the four-cell stage randomises left-right patterning. Frog. EST. PMID 16554361.

E3. Which end grows a head is a voltage decision. Block the H+/K+ pump at a flatworm's head-end wound and it regenerates a tail where a head belonged. Planaria. EST. PMID 21276941.

E4. A complete extra eye, built where no eye belongs. Hyperpolarising cells with Kir2.1 induces anatomically complete ectopic eyes in gut, tail and somite tissue, only in competent tissue. Voltage says which organ, not how to build one. Frog. EST; mammals EV. PMID 22159581.

E5. And the extra eye sees. Tadpoles with eyes on their tails learn visual tasks through them. Frog. EST. PMID 23447666.

E6. A voltage dye spots a tumour before it looks like one, and a channel blocks it with no DNA edit. Frog. EST; any human diagnostic EV. PMID 23471912.

E7. Healthy cells suppress a tumour from a distance, through gap junctions. Frog. EST. PMID 23196890.

E8. The gold-standard control: restore the voltage with a different channel and the defect is rescued, so it was the voltage, not the protein. Frog. EST. PMID 26864374.

E9. A two-headed worm that stays two-headed across repeated cuttings, with the genome untouched. Planaria. EST for the phenomenon; UR for where the memory is stored, because nobody has found it yet. PMID 28538159.

E10. Twenty-four hours of drug, eighteen months of leg. A wearable bioreactor delivered a five-drug cocktail to an adult frog's amputation site for one day; the frog grew a partial, functional hind limb over the following year and a half. Frog. EST; mammals EV; independent limb-scale replication still pending. PMID 35080969.

The three overclaims, tagged honestly

The anatomical compiler does not exist. Individual write primitives exist. EV.

Human limb regeneration has not been done. Adult frogs got a partial hind limb. EV.

There is no bioelectric cancer cure. Tumours were blocked in frog embryos. EV.

Overstating these is the most criticised failure of the field, and the reason a claim ladder is worth the discipline. A tag can go down; it can never be argued up.

Why this is the access layer

The genome is a parts list. The voltage pattern is closer to the address space: change it and competent tissue builds a different organ, with the same parts. It is readable with dyes and probes, writable with channels and drugs, and it acts at the millimetre scale where developmental decisions are made, spread cell to cell through gap junctions.

That is what a privileged access layer means in software: the place you can write once and have the rest of the system follow. It is why the second premise of the mission is not a hope.

genome (parts list)
      │
voltage pattern across tissue  ◄── read: dyes, probes   ◄── write: channels, drugs
      │
tissue builds: eye · head · tail · limb · (no tumour)

all writing