Text Optimize

sonnet session

Quick reference

Field	Value
Command	`/brewtools:text-optimize`
Arguments	`[-l\|-d] [file\|folder\|path1,path2]` — `-l` light, `-d` deep, no flag = medium
Model	sonnet
Context	session
Tools	Read, Write, Edit, Grep, Glob, Task

Modes

Mode	Flag	Best for	What changes
Light	`-l`	Critical files, production prompts	Filler removal, tone fixes, reference checks — structure untouched
Medium	(default)	General docs, agents, skills	Tables, bullets, merged sections, full rule set
Deep	`-d`	Large prompts where every token counts	Aggressive rephrasing, max compression — always review diff after

Usage examples

# Single file, medium mode (default)
/brewtools:text-optimize CLAUDE.md

# Light mode on a production agent -- safe, minimal changes
/brewtools:text-optimize -l .claude/agents/reviewer.md

# Deep mode on a verbose prompt you want compressed
/brewtools:text-optimize -d prompts/analysis-prompt.md

# Multiple files processed in parallel
/brewtools:text-optimize agents/planner.md, agents/executor.md, CLAUDE.md

# All markdown files in a directory
/brewtools:text-optimize -d agents/

What it does

Converts verbose prose to dense tables (up to 3x more token-efficient)
Removes filler words and passive constructions
Restructures numbered lists to bullets where order does not matter
Converts multi-line code blocks to inline code when a single expression suffices
Merges redundant or overlapping sections
Applies positive framing (“do Y” instead of “don’t do X”)
Verifies all file paths (R.1), URLs (R.2), and circular references (R.3)
Uses standard abbreviations in tables only (full words in prose)

Rule categories

Category	Rule IDs	Scope
Claude behavior	C.1-C.6	Literal following, avoid “think”, positive framing, match style, descriptive instructions, overengineering
Token efficiency	T.1-T.8	Tables, bullets, one-liners, inline code, abbreviations, filler, comma lists, arrows
Structure	S.1-S.8	XML tags, imperative, single source, context/motivation, blockquotes, progressive disclosure, consistent terminology, ref depth
Reference integrity	R.1-R.3	Verify file paths, check URLs, linearize circular refs
Perception	P.1-P.6	Examples near rules, hierarchy, bold keywords, standard symbols, instruction order, default over options

Mode-to-rules mapping

Mode	Applies	Notes
Light	C.1-C.6, T.6, R.1-R.3, P.1-P.4	Text cleanup only — no restructuring
Medium	All rules (C + T + S + R + P)	Balanced transformations
Deep	All rules + aggressive rephrasing	Merge sections, max compression

Output

Each file produces an optimization report:

Section	Contents
Metrics	Lines and tokens — before, after, percent reduction
Rules Applied	Which rule IDs were used and what changed
Issues Found	Broken references, redundancies, structural problems — and how they were fixed
Cross-Reference Check	Verification status for file paths, URLs, circular refs

Files are modified in-place. The report is printed to the conversation.

Tips

Start with light mode (-l) on important files to preview changes. Review deep mode diffs carefully — aggressive rephrasing can alter meaning in domain-specific content. Parallel processing kicks in automatically for multiple files or directories.