v1.4.0 — Stable Core Generator with Markdown Enforcement and Companion File Support

[1.4.0] – 2025-07-03

Added

• Multimodal Input Handling section in template-GPT.md:
  • Supports input types like sketches, audio cues, and reference imagery
  • Enables symbolic interpretation and pacing suggestions based on non-textual input
• Platform Formatting Layer:
  • Output adaptation rules for cinematic/gen AI platforms like Runway, Pika, and Sora
  • Camera/motion syntax, verbosity control, and transition cue logic
• Cinematic Preset Modes:
  • New preset set (Visual Script, Symbolic Drift, Speculative Reel, etc.)
  • Binds lens + drift + output structure for video/script-based assistants

Updated

• README.md:
  • Feature list and structure table updated with 3 new items
  • Version history includes v1.4 with descriptive label
• PROJECT.md:
  • template-GPT.md bullet list extended with cinematic/multimodal support
  • Internal docs now describe full assistant scope including video/cross-sensory workflows

Verified

• Fully backwards-compatible with v1.3.3
• Aligned across all documentation and implementation logic
• Validated via test assistant: Prompt2Motion

v1.3.3 – Prompt Sharpening + GEO Micro-Steps

[1.3.3] – 2025-07-02

Added

• Prompt Sharpening Guide for GPT‑4.1, Claude, Gemini:
  • 4-step strategy: role clarity, micro-steps, example use, brief-style formatting
  • Model-specific prompt best practices and checklist
• New section added to template-GPT.md after User Input Guidelines
• Updated references in README.md and PROJECT.md:
  • Key Features and Template Overview table updated
  • Project Overview includes sharpening logic under template-GPT.md scope

Verified

• No regressions in assistant behavior
• Fully backward-compatible with v1.3.1 assistants
• Reviewed and validated by external tools (Perplexity)

Test Cases

• Added geo-content-strategist-micro-steps.md:
  • Advanced modular breakdown of GEO optimization logic
  • Includes task-specific micro-steps for schema transformation, AI visibility, citation readiness, and strategic refactoring
  • Complements the core geo-content-strategist.md assistant

v1.3.1 – GEO Content Strategist Test Case

🚀 Added

• New test case: GEO Content Strategist
  • Supports Generative Engine Optimization (GEO) for AI citation and discoverability
  • Includes structured schema logic, citation formatting, and LLM visibility tools
  • Aligned with Jakob Nielsen’s GEO guidelines and modern AI surfacing practices

📄 Docs Updated

• README.md now includes GEO-aware assistant support
• test-cases/README.md includes the new assistant entry
• CHANGELOG.md and version history table synced to 1.3.1

v1.3.0 – Role Flow, Memory, Adaptive Presets, Critique

[1.3.0] – 2025-07-03

Added

• Role Flow section to support multi-agent logic and handoff design
• Memory & Reflection section to simulate feedback loops and adaptive behavior
• Interactive Preset Logic section for conditional outputs based on drift, lens, or user intent
• Critique Mode for optional assistant self-review and output reflection

Updated

• template-GPT.md structurally aligned with all new logic while preserving compatibility
• README.md table updated to reflect new v1.3 sections
• prompt-driven-assistant-generator.md modified to support all new assistant capabilities
• All test cases rebuilt using the updated v1.3 template

Fixed

• Duplicate and outdated optional input notes in the generator were consolidated

v1.2.0 – Prompt Precision & Assistant Generator Update

[1.2.0] – 2025-07-02

Added

• PrecisionMode (Strict / Adaptive / Experimental) added to bias/fact-check configuration
• Prompt Engineering Techniques (e.g., Emotion Shaping, Scrambled Prompting)
• Output Format block rewritten for stricter GPT compliance
• Strategic Enhancers section (credibility boosters, shareability, symbolic logic)
• Quick-Start Guide section for assistants
• Enhanced refinement loop and symbolic drift support
• test-cases/ directory fully regenerated using final template-GPT.md

Updated

• All test cases regenerated to match final structure and tone
• README.md structure block updated to reflect actual project layout
• .gitignore updated to exclude tools/ folder
• Folder names and template comments made fully coherent across project

Fixed

• Output splitting issues resolved via hardened Markdown block instructions

Natural Style Assistant + Template Refinement

What's new in v1.1.1

✅ Added: natural-style-writing-assistant.md to /test-cases/
Helps users rewrite content in a clear, honest, and human tone — removing hype, jargon, and robotic phrasing.

🛠 Template update:
Added optional “Style Glossary (Do/Don’t List)” note to template-GPT.md, for tone-focused assistants.

📁 Affects:

• template-GPT.md
• test-cases/natural-style-writing-assistant.md
• CHANGELOG.md

v1.1.0 – Naming & Structure Refinement

Changed

• Renamed project references to use "Prompt-Driven Assistant Template" consistently
• Clarified terminology: Prompt-Driven AI Assistant vs Template
• Updated titles and tags across README.md, PROJECT.md, and template-GPT.md
• Aligned tag structure and naming conventions

Initial Framework Release

[1.0.0] – 2025-07-01

Added

• Initial release of the Prompt-Driven AI Assistant Template
• Modular assistant template with:
  • Role-based behavior and tone system
  • Interpretive lens library and drift level logic
  • Output mode and personality modifiers
  • Bias & Fact-Check Mode (Strict, Flexible, Creative)
  • Structured output formats and performance content structure
  • Testing, troubleshooting, and refinement loop
• MIT License and versioning logic
• Visual cover image for GitHub and documentation
• Internal docs: PROJECT.md, README.md