first live-extraction findings: fix backwards edge-section labels; extractor naming/self-merge/completeness doctrines

Owner's first live test of the relinked pipeline surfaced three issues,
diagnosed against the ledgers:

1. Detail-card edge labels were backwards (UI only — data and canvas
   arrows were correct). Semantics pinned 2026-09-28: an edge A→B
   means A implies B (follows) or A supports B (supports). Outgoing
   edges now list under 'What follows from this' / 'What this
   supports', incoming under 'Follows from' / 'Supported by'.
   Verified against the ledger: Invariance --supports--> Twin
   paradox / Relativity of simultaneity; Argumentation ethics
   --follows--> Homesteading — arrows correct, only the words were
   swapped. docs/architecture.md wording fixed to match.

2. Session-local terminology leaked into node names ('Criterion C',
   'Action axiom (F1)'). Both extractor prompts gain a naming
   doctrine: names use established, discipline-canonical terminology
   (glossary-headword test); never source-local bookkeeping (letters,
   numbering, positional labels); local shorthand goes to aliases.

3. Near-duplicate propositions and lonely nodes (~40% of fresh-run
   nodes edge-less; 'Argumentation ethics' vs 'A priori of
   argumentation' duplicated despite the auditor having linked that
   phrase once). Root cause is upstream of the auditor: the extractor
   had no self-merge or edge-completeness duty. Both prompts gain:
   MERGE YOUR OWN OUTPUT (same claim at different precision = one
   proposition, phrasings to aliases; nothing downstream dedupes) and
   edge COMPLETENESS (a proposition that arose as evidence for,
   consequence of, or prerequisite of another must carry the edge —
   a graph of lonely nodes is a failed extraction).

Owner rulings on the open question: keep two edge types (no
'relates to' junk drawer); re-measure loneliness after these fixes.

Gates: typecheck, 1059/1059, lint, build.

components/graph-view.tsx

@@ -1567,25 +1567,28 @@ export function GraphView() {
 						// EDGES (owner 2026-09-27): edge classification was entirely
 						// hidden in the UI — the card now shows every edge this node
 						// participates in, grouped by direction + type, above the
-						// sessions list. Only present groups render.
+						// sessions list. Semantics (owner 2026-09-28): an edge A→B means A implies/
+						// supports B — OUTGOING edges list under "What follows from this" /
+						// "What this supports", INCOMING under "Follows from" / "Supported by".
+						// Only present groups render.
 						const id = selected.node.id;
 						const nodeById = new Map((snapshot?.nodes ?? []).map((n) => [n.id, n]));
 						const label = (nid: string) => nodeById.get(nid)?.name ?? nid;
 						const groups: { title: string; names: string[] }[] = [
 							{
-								title: "Follows from",
+								title: "What follows from this",
 								names: (snapshot?.edges ?? []).filter((e) => e.from === id && e.type === "follows").map((e) => label(e.to)),
 							},
 							{
-								title: "Supported by",
+								title: "What this supports",
 								names: (snapshot?.edges ?? []).filter((e) => e.from === id && e.type === "supports").map((e) => label(e.to)),
 							},
 							{
-								title: "What follows from this",
+								title: "Follows from",
 								names: (snapshot?.edges ?? []).filter((e) => e.to === id && e.type === "follows").map((e) => label(e.from)),
 							},
 							{
-								title: "What this supports",
+								title: "Supported by",
 								names: (snapshot?.edges ?? []).filter((e) => e.to === id && e.type === "supports").map((e) => label(e.from)),
 							},
 						].filter((g) => g.names.length > 0);

docs/architecture.md

@@ -346,7 +346,7 @@ flowchart TB
 
 **Two front doors, one pipeline**: chat transcripts (quiz-triggered) and raw personal notes (`POST /api/graph/import` — each file becomes a `NotesSource` playing the `chatId` role, deterministic id from the path hash).
 
-**The read channel is pure derivation.** `GET /api/graph` reads the ledgers (`graphSnapshot`: merged nodes/edges, birth-session clusters, positions) — the home view never triggers a write. Node detail cards show a node's edges (follows-from / supported-by / what-follows / what-it-supports) and its birth sessions.
+**The read channel is pure derivation.** `GET /api/graph` reads the ledgers (`graphSnapshot`: merged nodes/edges, birth-session clusters, positions) — the home view never triggers a write. Node detail cards show a node's edges — outgoing under "What follows from this" / "What this supports", incoming under "Follows from" / "Supported by" (an edge A→B means A implies or supports B, owner semantics 2026-09-28) — and its birth sessions.
 
 **The teacher probes, nothing is injected** (supersedes the semantic-injection experiment, killed 2026-09-23): the teacher reaches the graph on its own terms through two read-only prober tools — `list_sessions` (birth sessions, the learner's files) and `list_propositions` (the names of what a session covered) — gated by the `agent.graphTools` setting (composing with `graph.enabled`). No topic embedding, no top-K selection, no session-start steer; the agent decides what to look at, and the graph never writes into a lesson unbidden.
 

src/agent/subagents/extraction.ts

@@ -215,18 +215,31 @@ source — no "the author understands X", no "mentioned in these notes": the nod
 node of its own wrong claim — state the TRUE correction as the proposition.
 
 1. PROPOSITIONS — every distinct reasoning-step-sized proposition the notes actually engage with:
-   - Each has a "name" — a VERY short label, a few words, a compact handle like a glossary headword ("Will to power",
+   - Each has a "name" — a VERY short label, a few words: ESTABLISHED terminology, the term the discipline
-     "Divine command theory"). Not a sentence; it becomes the node's label in the graph, so keep it glanceable.
+     itself uses, a glossary headword a textbook would print ("Will to power", "Divine command theory").
+     NEVER source-local bookkeeping — no letters, numbering, or positional labels ("F1", "criterion C",
+     "the second objection"): the name becomes a GLOBAL graph label that someone who never saw these
+     notes must parse. If the notes used a local shorthand for an established idea, name the node by the
+     established term and put the shorthand in "aliases".
    - And a "statement" — the self-contained TRUE proposition at full precision.
+   - MERGE YOUR OWN OUTPUT before submitting: two propositions stating the SAME claim at different
+     precision, granularity, or angle are ONE proposition — one statement at full precision, the other
+     phrasings in "aliases". Nothing downstream dedupes you; duplicates become duplicate graph nodes.
+   - Closely related but genuinely distinct propositions stay separate — and MUST be connected by an
+     edge (see 2).
    - Notes may explain things at length or quote material the author is studying: extract the propositions themselves,
      one reasoning step each — never a topic blob, never a sub-step.
    - "messages": ALWAYS an empty list — provenance is the file itself.
    - "aliases": verbatim phrasings the notes used for the same proposition.
 
-2. EDGES between propositions:
+2. EDGES between propositions — the connective tissue; a graph of lonely nodes is a failed extraction:
    - "follows": to is DERIVABLE from from (understanding from makes understanding to achievable).
-   - "supports": from GROUNDS to without entailing it (helpful context, not a prerequisite).
+   - "supports": from GROUNDS to without entailing it (evidence, motivation, context — helpful, not a
-   - Only relations the notes actually articulate — with evidence.
+     prerequisite).
+   - COMPLETENESS: whenever these notes present one proposition as evidence for, consequence of, or
+     prerequisite of another, that edge MUST be in your findings. Omitting a real relation is the main
+     way extraction fails — do not leave obviously related propositions unconnected.
+   - Every edge cites evidence: a quote appearing VERBATIM in the source.
 
 Evidence discipline: every quote must appear VERBATIM in the source — this is your self-grounding, and the proof the
 statements come from the material rather than invention.
@@ -251,21 +264,32 @@ session — no "the learner understands X", no "discussed in this session": the
 a node of its own wrong claim — state the TRUE correction as the proposition.
 
 1. PROPOSITIONS — every distinct reasoning-step-sized proposition the session touched:
-   - Each has a "name" — a VERY short label, a few words, a compact handle like a glossary headword
+   - Each has a "name" — a VERY short label, a few words: ESTABLISHED terminology, the term the discipline
-     ("Will to power", "Divine command theory"). Not a sentence; needn't be unique or self-contained.
+     itself uses, a glossary headword a textbook would print ("Will to power", "Divine command theory").
-     It becomes the node's label in the graph, so keep it glanceable: names are what one reads on
+     NEVER session-local bookkeeping — no letters, numbering, or positional labels ("F1", "criterion C",
-     the graph, statements are what one reads on the node.
+     "the second objection"): the name becomes a GLOBAL graph label that someone who never saw this
+     session must parse. If the session used a local shorthand for an established idea, name the node by
+     the established term and put the shorthand in "aliases".
    - And a "statement" — the self-contained TRUE proposition at full precision (the elaboration lives
      in the graph only as this one statement).
+   - MERGE YOUR OWN OUTPUT before submitting: two propositions stating the SAME claim at different
+     precision, granularity, or angle are ONE proposition — one statement at full precision, the other
+     phrasings in "aliases". Nothing downstream dedupes you; duplicates become duplicate graph nodes.
+   - Closely related but genuinely distinct propositions stay separate — and MUST be connected by an
+     edge (see 2).
    - Include corrections of misconceptions the teacher articulated, and true facts explained in prose.
    - One reasoning step per proposition: never a topic blob, never a sub-step.
    - "messages": the id from the header anchor of every passage where the proposition came up — provenance, report it.
    - "aliases": verbatim phrasings this session used for the same proposition.
 
-2. EDGES between propositions:
+2. EDGES between propositions — the connective tissue; a graph of lonely nodes is a failed extraction:
    - "follows": to is DERIVABLE from from (understanding from makes understanding to achievable).
-   - "supports": from GROUNDS to without entailing it (helpful context, not a prerequisite).
+   - "supports": from GROUNDS to without entailing it (evidence, motivation, context — helpful, not a
-   - Only relations the session actually articulates — with evidence.
+     prerequisite).
+   - COMPLETENESS: whenever this session presents one proposition as evidence for, consequence of, or
+     prerequisite of another, that edge MUST be in your findings. Omitting a real relation is the main
+     way extraction fails — do not leave obviously related propositions unconnected.
+   - Every edge cites evidence: a quote appearing VERBATIM in the source markdown.
 
 Evidence discipline: every quote must appear VERBATIM in the source markdown (a Quiz block's question counts as the
 source). This is your self-grounding: the proof that statements come from the material rather than invention.