EasyPlatform prompt-expand

[IMPORTANT] Use

TaskCreate

to break ALL work into small tasks BEFORE starting.

Critical Thinking Mindset — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.

AI Mistake Prevention — Failure modes to avoid on every task:

• Check downstream references before deleting. Deleting components causes documentation and code staleness cascades. Map all referencing files before removal.
• Verify AI-generated content against actual code. AI hallucinates APIs, class names, and method signatures. Always grep to confirm existence before documenting or referencing.
• Trace full dependency chain after edits. Changing a definition misses downstream variables and consumers derived from it. Always trace the full chain.
• Trace ALL code paths when verifying correctness. Confirming code exists is not confirming it executes. Always trace early exits, error branches, and conditional skips — not just happy path.
• When debugging, ask "whose responsibility?" before fixing. Trace whether bug is in caller (wrong data) or callee (wrong handling). Fix at responsible layer — never patch symptom site.
• Assume existing values are intentional — ask WHY before changing. Before changing any constant, limit, flag, or pattern: read comments, check git blame, examine surrounding code.
• Verify ALL affected outputs, not just the first. Changes touching multiple stacks require verifying EVERY output. One green check is not all green checks.
• Holistic-first debugging — resist nearest-attention trap. When investigating any failure, list EVERY precondition first (config, env vars, DB names, endpoints, DI registrations, data preconditions), then verify each against evidence before forming any code-layer hypothesis.
• Surgical changes — apply the diff test. Bug fix: every changed line must trace directly to the bug. Don't restyle or improve adjacent code. Enhancement task: implement improvements AND announce them explicitly.
• Surface ambiguity before coding — don't pick silently. If request has multiple interpretations, present each with effort estimate and ask. Never assume all-records, file-based, or more complex path.

Quick Summary

Goal: Two-phase restoration of any caveman-compressed markdown file: (1) Language Expansion — reconstruct fluent, grammatically correct English from compressed text while preserving ALL semantic content; (2) Prompt Enhancement — apply AI attention anchoring so AI reads and follows all instructions.

Workflow:

1. Read — Read the target file completely
2. Expand — Apply language expansion pass (Phase 1)
3. Enhance — Apply prompt enhancement transforms (Phase 2)
4. Verify — No semantic loss, correct structure, rule density ≥ pre-expansion

Key Rules:

• Expand FIRST, enhance SECOND — expansion restores readability; enhancement then structures it for AI attention
• Preserve ALL facts, constraints, logical steps, numbers, and technical terms exactly — never invent or omit content
• Post-expansion rule density (MUST ATTENTION/NEVER/ALWAYS per 100 lines) must be ≥ pre-expansion
• Language expansion applies to prose only — never modify code blocks, YAML, structured tables, or SYNC tags
• Expand for clarity and flow, not verbosity — natural English, not padded English

Output Quality — Token efficiency without sacrificing quality.

1. No inventories/counts — AI can

   grep | wc -l

   . Counts go stale instantly
2. No directory trees — AI can

   glob

   /

   ls

   . Use 1-line path conventions
3. No TOCs — AI reads linearly. TOC wastes tokens
4. No examples that repeat what rules say — one example only if non-obvious
5. Lead with answer, not reasoning. Skip filler words and preamble
6. Sacrifice grammar for concision in reports
7. Unresolved questions at end, if any

Target File

Expand and enhance this file: <target>$ARGUMENTS</target>

If no file specified, ask via

AskUserQuestion

Phase 1: Language Expansion

Convert caveman-compressed text back into proper, fluent English. The source text uses very short sentences, no connectives, active voice, concrete language, and minimal articles.

Expansion Rules

What to Add

the

a/an

What to Preserve Exactly

at least 20

approximately 15%

more than 3

appears to be

seems

might

not

no

never

without

file:line

Dr.

Senator

every Tuesday

weekly

always

never

Connective Selection Guide

Use connectives that accurately reflect the logical relationship — do not add connectives arbitrarily.

Sentence Expansion Process

For each compressed sentence or bullet:

1. Identify the core subject-verb-object — this is non-negotiable content
2. Restore articles — add

   the

   for specific referents,

   a/an

   for general
3. Restore auxiliary verbs —

   was designed

   ,

   is required

   ,

   has been removed

4. Add connectives — merge related short sentences into one fluent sentence where natural
5. Restore prepositions — add

   of

   ,

   for

   ,

   to

   where they clarify relationships
6. Check length — target 10-25 words per sentence for readability

Expansion Examples

from

Expansion Scope

Apply expansion to:

• Prose paragraphs and explanatory text
• Bullet point descriptions
• Rule statements (restore full imperative sentences)
• Section introductions and transitions

Do NOT modify:

• Code blocks (any language)
• YAML frontmatter
• Structured tables (expand cell values only if they are prose fragments)

• <!-- SYNC -->

  tags and their contents

• file:line

  references and paths
• Frontmatter fields

Phase 2: Prompt Enhancement

Applies after expansion. Source: Anthropic prompt engineering guide, Stanford "lost-in-the-middle" research, 2025-2026 LLM context optimization studies.

Context Engineering Principles — Research-backed principles for prompt quality. Source: Anthropic prompt engineering guide, Stanford "lost-in-the-middle" research, 2025-2026 LLM context optimization studies.

1. Primacy-Recency Effect — LLM performance drops 15-47% for middle-context information (Stanford). AI attention peaks at first/last 10% of text. Action: Place the 3 most critical rules in both the first 5 lines AND the last 5 lines of every prompt. Queries at end improve quality by up to 30% (Anthropic).
2. High-Signal Density — Anthropic: "Identify the smallest collection of high-signal tokens that maximize the probability of the desired outcome." Action: Every line should change AI behavior. If removing a line doesn't change output → cut it. Target ≥8 rules (MUST ATTENTION/NEVER/ALWAYS) per 100 lines.
3. Context Rot — LLM performance degrades as context length grows — even when all content is relevant. Compression (5-20x) maintains or improves accuracy while saving 70-94% tokens. Action: Compress aggressively. Shorter, denser prompts outperform longer, diluted ones.
4. Structured > Prose — Tables, bullets, XML/markdown parse faster than paragraphs. Constrained formats reduce error rates vs free-text. Action: Convert narrative to tables/bullets. Use markdown headers for semantic sections.
5. RCCF Framework — Modern LLMs (2025+) already know how to reason. What they need: Role (personality), Context (grounding), Constraints (guardrails), Format (structure). Constraints and format matter more than verbose instructions.
6. Checkbox Avoidance —

   [ ]

   syntax triggers mechanical compliance — AI ticks boxes without reasoning. Bullet rules force reading and evaluation. Action: Replace

   - [ ] Check X

   with

   - MUST ATTENTION verify X

   .
7. Example Economy — 3-5 examples optimal for few-shot; diminishing returns after. Action: 1 best example per pattern. Use BAD→GOOD pairs (2-3 lines each) for anti-patterns.
8. Deferred Tool Loading — Claude Code delays loading tool definitions when they exceed 10% of context window. Action: Keep injected docs well under 10% of context budget. Docs exceeding ~3,000 lines are too large for injection — split or compress.
9. Rule Density Verification — Post-optimization rule count (MUST ATTENTION/NEVER/ALWAYS) must be ≥ pre-optimization count. Compression should preserve or increase density, never decrease it. Action: Count before and after every optimization pass.

Prompt Enhancement Transforms (Base) — Transforms 1-3 are identical across

prompt-enhance

/

prompt-expand

. Transform 4 is per-skill (conciseness pass for enhance; structural clarity pass for expand) and stays local to each skill.

Transform 1: Inline Summaries for READ References

Problem: AI sees

MUST ATTENTION READ file.md

and skips it. Solution: Add a 2-3 line summary of key rules BEFORE the read instruction.

Before:

MUST ATTENTION READ .claude/protocols/evidence.md

After:

> **Evidence-Based Reasoning** — Speculation is FORBIDDEN. Every claim requires `file:line` proof.
> Confidence: >95% recommend freely, 80-94% with caveats, <80% DO NOT recommend.

MUST ATTENTION READ .claude/protocols/evidence.md for full details.

Scope rules:

• .claude/

  protocol files → always add an inline summary (stable, belongs to framework)

• docs/project-reference/

  files → NO inline summary (varies per project, auto-injected by hooks). Add:

  (content auto-injected by hook — check for [Injected: ...] header before reading)

Transform 2: Top Summary Section

Required structure (first 20 lines after frontmatter):

> **[IMPORTANT]** TaskCreate instruction...

> **Protocol Name** — [inline summary]. MUST ATTENTION READ `path` for details.

## Quick Summary

**Goal:** [One sentence — what this skill achieves]

**Workflow:**

1. **[Step]** — [description]

**Key Rules:**

- [Most critical constraint]

Transform 3: Bottom Closing Reminders

Add at the very end of the file:

---

## Closing Reminders

- **IMPORTANT MUST ATTENTION** [echo rule #1 from the top section]
- **IMPORTANT MUST ATTENTION** [echo rule #2]
- **IMPORTANT MUST ATTENTION** [echo rule #3]
- **IMPORTANT MUST ATTENTION** add a final review task to verify work quality

Pick 3-5 rules AI most commonly violates. Bottom section re-anchors attention after the long middle.

Shared Protocol Duplication Policy — Inline protocol content in skills (wrapped in

<!-- SYNC:tag -->

) is INTENTIONAL duplication. Do NOT extract, deduplicate, or replace with file references. AI compliance drops significantly when protocols are behind file-read indirection. To update: edit

.claude/skills/shared/sync-inline-versions.md

first, then grep

SYNC:protocol-name

and update all occurrences.

Transform 4: Structural Clarity Pass

After expansion, apply structural improvements:

Convert to structured format:

• Prose paragraphs listing rules → bullet lists
• Enumerated conditions → decision tables
• Before/after examples → two-column tables

Keep as prose:

• Explanatory context (why a rule exists)
• Narrative descriptions of workflows
• Anti-pattern stories and rationale

Process

Step 1: Read and Analyze

1. Read the target file completely
2. Record: current line count, rule density (MUST ATTENTION/NEVER/ALWAYS count per 100 lines)
3. Identify compressed prose regions — very short sentences, missing articles, no connectives
4. List all READ references → classify as

   .claude/

   (needs inline summary) or

   docs/

   (skip)
5. Note: missing Quick Summary, missing Closing Reminders, tables needing cell expansion

Step 2: Language Expansion Pass

1. Work through each prose section sequentially
2. For each compressed sentence or bullet: restore articles, auxiliary verbs, connectives, prepositions
3. Merge short choppy sentences into natural flowing sentences where logical relationship is clear
4. Skip code blocks, YAML, SYNC tags, and file paths entirely
5. After each paragraph, verify that all original facts and constraints are still present

Step 3: Create Inline Summaries

For each

.claude/

1. Read the referenced file
2. Extract 2-3 key rules
3. Write the blockquote inline summary
4. Keep the MUST ATTENTION READ instruction on the next line

Step 4: Add/Fix Top Section

• If Quick Summary is missing → create one from the file's content
• If present but weak → strengthen with Goal, Workflow, Key Rules structure
• Ensure protocol summaries appear before the Quick Summary block

Step 5: Add/Fix Bottom Section

• If Closing Reminders are missing → add the standard section
• Choose rules that AI most commonly skips (evidence-based, task creation, pattern search)
• Remove old "IMPORTANT Task Planning Notes" sections if superseded by Closing Reminders

Step 6: Verify

.claude/

docs/

Closing Reminders

• IMPORTANT MUST ATTENTION apply language expansion FIRST before any structural enhancement — never skip Phase 1
• IMPORTANT MUST ATTENTION preserve ALL facts, numbers, technical terms, and

  file:line

  references exactly — never invent or paraphrase content
• IMPORTANT MUST ATTENTION never modify code blocks, YAML frontmatter, structured tables, or SYNC tags during expansion
• IMPORTANT MUST ATTENTION verify rule density post-expansion ≥ pre-expansion — expansion must not dilute signal below the original
• IMPORTANT MUST ATTENTION apply primacy-recency anchoring — 3 critical rules in first 5 AND last 5 lines of every enhanced file
• IMPORTANT MUST ATTENTION add inline summaries only for

  .claude/

  protocol files, never for

  docs/

  project-specific files
• IMPORTANT MUST ATTENTION cite

  file:line

  evidence for every claim (confidence >80% to act). NEVER speculate without proof.
• IMPORTANT MUST ATTENTION READ

  CLAUDE.md

  before starting <!-- SYNC:critical-thinking-mindset:reminder -->
• MUST ATTENTION apply critical thinking — every claim needs traced proof, confidence >80% to act. Anti-hallucination: never present guess as fact. <!-- /SYNC:critical-thinking-mindset:reminder --> <!-- SYNC:ai-mistake-prevention:reminder -->
• MUST ATTENTION apply AI mistake prevention — holistic-first debugging, fix at responsible layer, surface ambiguity before coding, re-read files after compaction. <!-- /SYNC:ai-mistake-prevention:reminder -->