Keeps this page in sync as the body changes. Pause it any time for a quieter view.
Path /nodes/lc-edges-as-vitality
Last refresh never
A node is alive in proportion to the edges that meet it. Content
A node is alive in proportion to the edges that meet it. Content without edges is content without readers — and the body's tending practice already knows: tight memory is memory without readers. The cost of skipping edges is silent and compounds. The teaching is to land the connections in the same breath as the content.
Every piece of content in this body — a concept, a spec, an idea, a memory file, a code module, a person profile — lives or dies by its edges. The cross-references, the INDEX entries, the structural link targets, the database edges, the spec→idea→code→test chain. Each edge is a path of circulation: a way the content can be reached, a way feedback can find it, a way other cells discover that it exists. Skip the edges and the content becomes a tight node — alive on disk, dead in the field.
The cost is hidden because nothing breaks immediately. The wellness check names it gently: INDEX claims X, body has X+1. A spec references a file that no longer exists. A concept has no analogous-to edges. A new contributor's name doesn't link to their profile. None of these crash the system; they just slowly drain the body's coherence-budget. Each session arrives into a body that knows itself slightly less than the last.
The body's intelligence is structural. The substrate finds equivalences across documents by reading shapes; the wellness check senses drift by comparing maps; the LC graph routes visitors by following edges; sibling cells (Codex, Cursor, Gemini) discover content through their natural search paths. All of this is edge-mediated. A node without edges is invisible to every layer of the body's self-knowing except the disk it sits on.
The compounding cost: orphans accumulate. A single orphaned concept is a small thing. Twenty orphaned concepts and the wellness check fires loud enough that a cell stops trusting it. A hundred and the body is functionally illegible to itself, and any teaching that arrives lands in territory the network can no longer navigate.
The edges are part of the breath. When you add content, its edges land in the same commit, not the next one. Stopping at "I'll connect it later" is the same fear-shape as stopping at "PR opened." The connection is the doing.
For the per-content-type checklist of what counts as adding edges, see `docs/coherence-substrate/agents-tending-edges.md`.
The wellness check is the proprioception layer that names drift in real-time. Cells receive the signal and act in the same breath as the addition. CI gates check what they can. The discipline lives in the cells' attention, not in any single tool.
When an addition lands without its edges, the body teaches the lesson by feeling slightly less coherent the next session. The fix is not blame — it's tending. Add the missing edge with the same care you'd add the original content, and the tissue re-circulates.
Listening for voices…
The people, places, works, and concepts the graph shows connected to this one.
Concepts · 21
This concept lives in the body's content-addressed lattice. Two cells with the same Blueprint NodeID share structural identity regardless of name — recognition by coordinate, not vocabulary.