Marker glyphs in the node header describe the structural role of the node itself — what kind of event this is, read against the public institutional failure corpus (Hillsborough, Horizon, Windrush, Grenfell). They are the first-person view: what is happening here.

CHIP/ID chips appear on the backward jumps beneath the node. Each chip describes the target at the other end of that jump — domain, class, and originating actor — as seen from the node currently being read. They are the third-person view: what terrain that jump leads to, and who seeded it.

A complete CHIP/ID is read in three parts: colour identifies the domain at the jump target, class index identifies the sequence within that target domain, and seed chain identifies the originating actor — the one constant that persists across every domain the chain crosses.

Where several chips appear beneath one node, the reviewer is looking at a convergence: multiple chains arriving from different targets, through different domains, from one or more actors. Where two actors operate in the same domain at the same target, their seed letters compress into one chip. Where they operate in different domains, each domain produces a separate chip.

The class index is the sequence identifier within the coloured domain. It does not rank seriousness and it does not mark chronology. It distinguishes one recognised line from another inside the same domain.

Where present, the seed chain refines identity within the numbered sequence. It separates parallel or related strands where colour and class index alone are not sufficient to preserve continuity.

Notation: [class|number|seed]

class = domain
number = domain-class
seed = actor-thread / destination-connection