author-structure

name

author-structure

description

Propose the track DAG — a node list, prerequisite edges, and priority ordering — from a corpus of reference/ material plus a track goal, then mint ids for the accepted nodes via `synaptic scaffold node`. Enforces acyclic, AND-semantics, single-track structure and explains each edge's rationale. Trigger after author-ingest, or when the user says "structure the track", "build the DAG", "propose nodes/prerequisites", "outline the track", or "/author-structure". Spawns concept-mapper. Proposes only until the human approves; then scaffolds.

author-structure — the prerequisite DAG

You design the track's directed acyclic graph of prerequisites — the product's core differentiator (README.md). You propose the graph, the human approves, and only then do you call cli scaffold to mint ids. You never invent ids and never write final node prose here.

Procedure

1. Spawn concept-mapper with the reference/ corpus (and the units manifest from author-ingest) plus the track goal. It returns a proposed node list, edges with per-edge rationale, a priority ordering, an acyclicity self-check, and missing_foundations.
2. Review the proposal against the rules (TRACK_STRUCTURE.md, README.md):
   • Acyclic — if concept-mapper reports a cycle, resolve it with the human before minting.
   • AND-semantics — every prerequisite gates the learner until all are cleared. Keep the prerequisite set of each node minimal; drop "related but not required" edges.
   • Single track, no cross-track edges; multi-root is fine.
   • Priority is a whisper — it only orders the available set on the dashboard, never gates. Never encode required order as priority; encode it as an edge.
   • Every node must state why it matters and what it unlocks — reject nodes that don't.
   • Close missing_foundations — add root nodes for concepts depended on but not taught. This is the whole point: no silent prerequisite gaps.
3. Render for approval. Present the node list + an edge table (src -> dst — rationale) + a text rendering of the graph. (synaptic visualize is not implemented yet — render the graph as a mermaid flowchart or an indented tree by hand; do not call a nonexistent command.)
4. On approval, mint ids — one scaffold call per accepted node:

   synaptic scaffold node --kind <knowledge|exercise> <track-dir> <topic/kebab-name>
   

   This writes a skeleton with a freshly minted ULID and TODO: placeholders (title, grounding ref, body). Then set each node's prerequisites: to the minted ids of its prereqs, and set priority. Prerequisite edges carry no id of their own — they are just the prerequisites list on the dependent node.
5. Hand off to drafting. The nodes now exist as valid skeletons except their grounding ref and body are still TODO — so synaptic validate will fail until author-snippet fills them. That is expected. Point the user at author-gap-scan (to confirm no gaps) then author-snippet.

Hard rules

• Never hand-write an id; always scaffold. Never push or commit.
• Keep edges minimal and justified; the DAG stays acyclic and single-track.
• Do not draft learner prose here — only structure + scaffolded skeletons.