System 1
Anthropic (Claude Guidelines)
Claude Native
Purpose: Built specifically for the Anthropic ecosystem. Uses native XML and the new Opus 4.7 Framework for maximum precision.
Result
✦
Prompt Optimizer
Static Mode
~/Desktop/AI/PROMPTS
Professional Suite
Unified Prompt Engineering.
Task Mapping & Strategy
Describe your ultimate goal below. Our logic will recommend the optimal engine and model stack for the build.
Engineering Terminals
Select an engine to begin the transformation
System 2
Open AI Vibe Coding
Vibe Coding
Purpose: The most powerful prompt for GPT-5.5 to build production-grade apps end-to-end under 5 minutes.
Hold Cmd/Ctrl to select multiple
Vibe Prompt
System 3
General LLM Optimizer
Universal Catalyst
Purpose: Cross-platform engine optimizing for general LLMs (GPT-4, Gemini, Kimi, etc.).
Result
System 4
Quick Optimization
Fast Logic
Formula: “I want you to act as a [ROLE] to [TASK] for [GOAL], using [CONTEXT], following [CONSTRAINTS], and ask clarifying questions before responding in [OUTPUT FORMAT].”
System Instruction Formula:
“I want you to act as a [ROLE] to [TASK] for [GOAL], using [CONTEXT], following [CONSTRAINTS], and ask clarifying questions before responding in [OUTPUT FORMAT].”
Refined Variants
System 5
JSON Prompt Optimizer
Prompt as Code
JSON System: Use this when the goal is to get the AI to output structured data for your apps or coding agents.
Verdict: Yes, keep them separate. JSON is for the output; XML is for the input architecture.
JSON SYSTEM NOTE:
Use this when the goal is to get the AI to output structured data for your apps or coding agents. JSON is for the output; XML is for the input architecture.
JSON Payload
System 6
XML
xml structuring
XML System: Use this when the goal is to structure the prompt itself to get better reasoning and instruction-following from the model.
Verdict: Yes, keep them separate. JSON is for the output; XML is for the input architecture.
XML SYSTEM NOTE:
Use this when the goal is to structure the prompt itself to get better reasoning and instruction-following from the model. JSON is for the output; XML is for the input architecture.
xml result
System 7
Markdown
Formatting Expert
Purpose: Converts any raw text, data, or transcript into perfectly structured GitHub-flavored Markdown. Ideal for documentation, READMEs, and technical notes.
Markdown Result
System 8
Karpathy Autoresearch
Iterative Refinement
Purpose: Peer-reviewed optimization based on Andrej Karpathy's methodology. Try small changes, measure, keep or revert.
Research Agent Prompt
System 9
Anti AI Writing
Human-Centric
Methodology: High-judgment writing rules designed to strip away "AI-isms". Focuses on clarity, specificity, and human rhythm while banning common LLM patterns and puffery.
Refined Writing Prompt
System 10
Prompt Eval Loop
Self-Improving
Methodology: Arise/Aakash eval loop. Score your prompt against test cases, identify failure patterns, and generate an improved version automatically.
Eval Pipeline:
Prompt + Test Cases → Score → Failure Patterns → Improved Prompt → Iterate
Eval Report + Improved Prompt
/goal
/goal Mega Prompt
Agentic Mode
Purpose: Builds a complete /goal mega-prompt for Claude Code, Codex, ChatGPT, and Hermes Agent. Drop it in and agents run for hours with zero hand-holding.
Drop into Claude Code or Codex or Hermes · agents run for hours · zero hand-holding
Begin by outputting your plan. Then execute end-to-end without checking in until done or genuinely blocked.
Set the goal · walk away · come back to shipped work
01
The Final Outcome
What "done" looks like in one line — this becomes your /goal
02
Context
Project · Stack · Current state · Working dir · Constraints · Audience
03
Success Criteria
All must be true. Be measurable. Final deliverable runs without errors.
1.
2.
3.
04
Operating Rules
Non-negotiable agent behaviors — toggle off to relax a rule
05
Quality Bar
06
Final Deliverable Checklist
Your /goal Mega Prompt
Prompt History
Save and manage your generated prompts for future reference and iteration
No history yet. Generate a prompt to get started.