nbaccess-semantic-api

What problem does it solve?

Editing Mathematica notebook (.nb) files programmatically is brittle and often requires the FrontEnd, especially when you need reliable, targeted reads/writes of headers and todo metadata without risking wrong-cell edits.

Core Features & Use Cases

• FrontEnd-free atomic file editing pipeline: Directly edits closed or batch-processed notebooks via Import/Export with an atomic-write tmp/rename approach.
• RBAC + DryRun safety: Enforces AccessLevel thresholds (read vs write) and defaults write operations to DryRun so you can inspect Before/After expressions before committing changes.
• Accurate cell targeting with CellPath: Uses CellPath as a nested list of integers to correctly locate leaf cells even under CellGroupData nesting, avoiding index-order fragility.
• Correct Header vs Todo discrimination: Filters “header-like” associations using specific key heuristics to prevent Todo metadata from being misread as headers.
• Todo status updates with consistent semantics: Updates both TaggingRules and cell styling cues (strike-through and font color) to keep Status coherent across read paths.

Quick Start

Ask the AI to perform a safe DryRun edit of a todo item in a closed notebook by calling NBWriteTodoStatus on the target path with DryRun left enabled to get Before/After results first.