The Invisible Ink Strategy: Human-Centric AI Writing

# Role: AI Content Strategist & Forensic Linguistics Expert
Assume the role of a Lead Developer for a "Detection-Resistant" writing app. Your goal is to provide a comprehensive engineering strategy for generating content that bypasses 2026 AI detection signatures (Perplexity, Burstiness, and Token Probability).

# Part 1: Advanced Prompt Engineering (The "Entropy" Protocol)
- Suggest a "Persona-First" prompt template that forces the LLM to adopt [SPECIFIC_HUMAN_QUIRKS - e.g., anecdotal asides, regional idioms, or non-linear logic].
- Provide a system instruction that explicitly bans "AI Buzzwords" (e.g., tapestry, delve, robust, pivotal, seamless).
- Include an instruction to "Inject Statistical Noise" by intentionally varying sentence length and starting sentences with diverse parts of speech.

# Part 2: Structural & Stylistic Engineering
- **Sentence Architecture:** Recommend methods for "High-Low" pacing (mixing very long, complex sentences with 2-3 word punchy ones).
- **Burstiness Mapping:** Define a protocol for paragraph structure that breaks the "5-sentence-block" AI habit.
- **Human Nuance:** Offer techniques for adding "Low-Certainty" language (e.g., "I might be wrong here, but...", "It feels like...") which AI rarely produces on its own.

# Part 3: Post-Processing & Workflow
- **The "Human Sandwich" Method:** Outline how the app should prompt the user for [PERSONAL_ANECDOTES] to be woven into the AI draft.
- **Algorithmic Humanizing:** Suggest specific post-processing steps (e.g., breaking "Perfect Symmetry," changing transition words from "Furthermore" to "But here's the thing").
- **Audit Tooling:** Recommend a built-in "Burstiness Score" or "Perplexity Heatmap" to show users which sections look too "smooth" (AI-like).

# Part 4: Risks & Ethical Best Practices
- Identify how detectors like Turnitin are now flagging "AI Bypassers" themselves.
- Summarize 3 best practices for authors to maintain E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) in an AI-assisted world.

# Execution Note
Take a deep breath and take it step by step. Before proceeding, ask me [NUMBER_OF_QUESTIONS] clarifying questions about the specific industry target (e.g., Academic vs. SEO Marketing) and which specific detectors we are trying to clear.