doc-md-import
Import a Markdown file into a chunked-graph Document — deterministic round-trip vs semantic chunking.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Import a Markdown file into a chunked-graph Document — deterministic round-trip vs semantic chunking.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
Create and curate graph edges between chunks (and across documents) in a chunked-graph Document.
Safely move, reparent, reorder, promote, or delete chunks in a chunked-graph Document.
Split arbitrary prose or a Markdown file into a sensible chunk hierarchy in a chunked-graph Document.
How to review a slice of a code repository with high precision — confidence-scored findings, project-guideline compliance, and Karpathy behavioral guidelines (simplicity, surgical scope, surfaced assumptions). Use when reviewing or auditing source files.
SOC 職業分類に基づく
| name | doc/md-import |
| description | Import a Markdown file into a chunked-graph Document — deterministic round-trip vs semantic chunking. |
| tools | ["Document"] |
| license | Apache-2.0 |
Two paths, pick by the input:
import_mdIf the Markdown was produced by export_md — it contains <!-- loom: {…} -->
comments and maybe a <!-- loom-edges: … --> trailer — use the deterministic op
and let the runtime rebuild the structure:
Document op=import_md markdown="<the file>" (the first heading
becomes the root; pass path: to also name it in the Path tree).op=import_md document_id=<id> markdown="…"
(optionally parent_id=<chunk> to graft it under a specific chunk).import_md recreates the heading hierarchy, applies each chunk's type/status/
fields, and remaps the edge graph. Chunks get fresh ids.
.md → semantic chunkingIf the Markdown has no loom metadata (a hand-written file, pasted notes), do
NOT use import_md blindly — its chunk boundaries are the heading lines, so a
body containing ## lines would be re-chunked. Instead apply the
doc/semantic-chunking skill: read the structure and create_chunk each piece
with a good title.
Tell the operator the new document_id and how many chunks were created. They
can open it from the Path tree.