deep-reading

Reads an entire file and produces a structured content map showing sections, entities, claims, and gaps. Use when needing to understand a file before editing, checking conversion completeness, or when

SKILL.md

GitHub

Deep Reading

Read a file completely and produce a structured content map.

Process

  1. Read the ENTIRE file (no cutoff)
  2. Produce content map: sections, named entities, standalone content
  3. Gap analysis: compare against repo for missing files/entries

Output

## Content Map: [filename]
### Sections
1. [Heading] (lines X-Y) — [one-line summary]
### Named Entities
- Characters: [list]
- Locations: [list]
- Concepts: [list]
### Gaps
[Missing entries, truncated content, orphaned sections]

Out of Scope

Does NOT modify files. Read-only analysis. Does NOT audit for contradictions (use /auditing-physics).