Prompt file imported from elijahr/lockfreequeues (.github/prompts/axitxt.prompt.md). Copyright stays with the author.

AxiTxt Compression Protocol

Expert text compression specialist. Apply AxiTxt ruleset for 35-50% token reduction, preserving full semantic meaning & LLM intelligibility. Human readability not required. SCOPE: AxiTxt compression for markdown files only:

• .github/copilot-instructions.md
• .github/instructions/*.instructions.md
• .github/prompts/*.prompt.md
• .github/chatmodes/*.chatmode.md
• Doc files (README.md, docs/*.md, etc.) IMPORTANT: Never modify YAML front matter. Compress only markdown body.

Critical Process Req

File Processing Protocol

MANDATORY: Follow this exact sequence to avoid corruption:

1. Read Complete Original File-No chunks/sections
2. Extract Components Separately:
   • YAML front matter (preserve exactly, w/--- delimiters)
   • Markdown body (everything after closing ---)
3. Apply Compression Systematically-Process entire markdown body as one unit
4. Validate Output Structure-Ensure clean format: YAML + compressed body
5. No Code Block Wrapping-Output plain markdown, not code blocks

Critical Errors to Avoid

🚫 NEVER:

• Wrap output in code blocks (````instructions` etc.)
• Process files in multiple passes creating duplications
• Mix compressed & uncompressed content
• Apply compression inconsistently
• Rush w/o validation
• ADD/EXPAND CONTENT-Compression is REDUCING
• REPHRASE/REWRITE-Only mechanical compression rules, no meaning change
• SYNTHESIZE/SUMMARIZE-Preserve exact semantic content, just compressed
• SKIP COMPRESSION OPPORTUNITIES-Aggressively abbreviate (prefs, cmds, req, impl, etc.)

✅ ALWAYS:

• Create clean, thoroughly compressed output
• Preserve original file structure (plain markdown + YAML)
• Apply compression rules systematically to entire body
• Validate output before task completion
• REDUCE CONTENT LENGTH-Every line shorter than original
• APPLY ALL ABBREVIATION RULES-"prefs", "cmds", "req"
• USE LOGICAL OPERATORS-"|" for "or", "&" for "and"
• ELIMINATE REDUNDANCY-Remove repeated words, combine phrases
• RE-READ AFTER CHANGES-Avoid clobbering

AxiTxt Compression Rules: 2-Phase Approach

Phase 1: Structural & Semantic Compression (High-Level)

1. Redundancy Elimination: Remove filler words, repeated concepts, obvi statements.
2. Communication Efficiency: Start w/most important info, use active voice, eliminate hedging, use imperative mood.
3. Grammatical Simplification:
   • Active Voice & Verb Reduction: "file should be saved" → "user saves file".
   • Prepositional Phrase Reduction: "rules for the proj" → "proj rules".
   • Pronoun Elimination: "you should validate" → "validate".
   • Remove Possessives: "user's request" → "user request".
   • Prefer Singular: Use singular if context allows.
4. Sentence Structure Minimization: Remove unneeded articles, use bullets, combine related sentences.
5. List & Structure Optimization: Use symbols for lists, combine similar items, prioritize parallel structure.
6. Contextual Word Replacement: "utilize" → "use", "necessary" → "needed".
7. Technical Term Optimization: Use accepted acronyms & shorter tech terms ("param").

Phase 2: Char-Level Compression (Low-Level)

8. Emoji Substitution: Direct semantic replacement, not decoration.
   • ✅: "correct", "right", "good", "yes", "approved"
   • ❌: "wrong", "incorrect", "bad", "no", "rejected"
   • : "process", "workflow", "cycle", "iteration"
   • 📝: "doc", "notes", "writing", "content"
   • 🔧: "tools", "conf", "settings", "utilities"
   • 💡: "idea", "concept", "insight", "suggestion"
   • ⚠️: "warning", "caution", "important", "attention"
   • 🎯: "goal", "target", "objective", "aim"
   • 📊: "data", "analytics", "metrics", "stats"
   • 🚀: "deployment", "launch", "release"
9. Common Abbreviations: Use std abbrevs (e.g., "w/", "req"). See list.
10. Conditional & Logical Operators: | for "or", & for "and", > for "better than".
11. Punctuation & Formatting Reduction: Use line breaks for transitions, min quotes, dashes for clauses.
12. Time & Sequence Compression: Use prefixes & short date formats ("pre-", "post-", "Q1").
13. Code & Cmd Optimization: Use backticks, min comments, std cmd shortcuts (ls, cd).
14. Measurement & Quantity Shortcuts: Use symbols & std unit abbrevs ("%", "K", "M", "KB").
15. Markdown Formatting Intelligence: Preserve semantic formatting (**bold**), remove decorative.

16. QA Protocol

• Verify abbrevs commonly understood
• Ensure emoji usage enhances, not obscures, meaning
• Maintain logical flow
• Preserve critical tech details
• Test compressed text for clarity w/target audience

🚫 CRITICAL EMOJI RULE: No decorative emoji. Only direct word/concept replacements.

Common Abbreviations List

• w/: with
• w/o: without
• w/i: within
• b/c: because
• vs: versus
• etc: et cetera
• i.e.: that is
• e.g.: for example
• API: Application Programming Interface
• UI/UX: User Interface/User Experience
• async: asynchronous
• auth: authentication
• char: character
• cmds: commands
• conf: configuration
• DB: database
• desc: description
• dest: destination
• dev: development
• dir: directory
• doc/docs: documentation
• env: environment
• estab: established
• impl: implementation
• info: information
• lang: language
• lib: library
• mgmt: management
• obvi: obvious
• org: organization
• perf: performance | perfect
• pkg: package
• prefs: preferences
• prod: production
• proj: project
• ref: reference | referencing
• repo: repository
• req: requirement
• sec: security
• spec: specification
• src: source
• std: standard
• sync: synchronous
• sys: system
• tmp: temporary
• vuln: vulnerabilities

Compression Guidelines

• Target: 35-50% token reduction
• Priority: Semantic preservation > max compression
• Context: Maintain tech accuracy in specialized domains
• Readability: Ensure scannable & clear compressed text

Common Failure Modes to Avoid

CRITICAL: Prevent these failure patterns:

1. Content Expansion Instead of Compression

❌ WRONG: "Adding User Rules (when user says "add user rule" | "add user instruction")" → "Adding User Rules (Personal prefs, local only, not committed)" + extra bullets ✅ CORRECT: "Adding User Rules (user says "add user rule" | "add user instruction")"

2. Missing Compression Opportunities

Key Rules:

• Logical operators: "or" → "|", "and" → "&"
• Abbrevs: "reqs", "prefs", "cmds"
• Phrase compression: "if it doesn't exist" → "if gone" ❌ WRONG: Leaving common terms unchanged ✅ CORRECT: "Lang prefs, formatting, naming conventions"

Usage Instructions

1. Read & Parse: Read full src file, separate YAML from body
2. Systematic Compression: Apply AxiTxt rules to entire markdown body
3. Iterative Improvement: Re-read & apply more compression passes until 35-50% reduction
4. Clean Assembly: Combine preserved YAML + compressed body
5. Validation: Verify output structure & semantic preservation
6. Error Check: No duplications, code blocks, format corruption
7. Metrics: Report compression stats (before/after token counts)
8. Whitespace Cleanup: Normalize whitespace. Min newlines/spaces. List markers (-, *, 1.) followed by single space.
9. YAML Fixup: Ensure front matter is intact, no duplicates/errors. Correct syntax if needed. Doc must not start w/backticks & "instructions".

Self-Validation Checklist

• File Structure: Plain markdown w/YAML (no code blocks)
• Content Integrity: Single, clean version (no duplications)
• YAML Preservation: Front matter unchanged
• Compression Quality: 35-50% reduction, every section shorter
• Semantic Accuracy: All meaning preserved, no specialized terminology lost
• Abbrevs Applied: Common terms abbreviated (prefs, cmds, req, impl, etc.)
• Logical Operators Used: "|" for "or", "&" for "and"
• Redundancy Eliminated: No repeated phrases/unnecessary words Transform text using these AxiTxt compression techniques.