Professional Prompt Intelligence
Find professional prompts and understand the AI products you use.
Find rights-reviewed prompts for professional work and learn how AI products handle instructions and workflows from published evidence.
Version evidence
Selected historical change
A selected comparison of November 2025 and January 2026 Claude prompt evidence.
- Claude: Product context and knowledge boundary expandedRead interpretation
Anthropic's highlighted January changes add named product surfaces and sharpen the instruction not to infer unlisted product details that may have changed.
Curated paths
Start with the work you need to do
Search 20 rights-reviewed prompts by task, domain, source, prompt type, or established target.
Compare curated agent pairs using their evidence, scope, and workflow context.
Select workflow requirements and inspect evidence-backed product fit across six products.
Read the evidence-backed interpretation, then inspect its reviewed semantic diff.
Product intelligence
Explore AI products and agents
Learn how six AI products handle instructions, context, memory, tools, permissions, and other workflow behavior.
| Product | Surface | Prompt intelligence | Review |
|---|---|---|---|
| OpenAI CodexOpen-source agentic coding interface | Codex CLI | Limited excerpt and analysis | Sep 05, 2026 |
| ClaudeWeb, desktop, and mobile AI assistant | claude.ai and Claude apps | Limited excerpt and analysis | Aug 11, 2026 |
| Claude CodeMulti-surface agentic coding product | Claude Code terminal CLI · Claude Code IDE integrations · Claude Code desktop app · Claude Code web | Instruction map; no full prompt | Aug 31, 2026 |
| Gemini CLIOpen-source agentic command-line interface | Gemini CLI | Documentation-derived analysis | Aug 11, 2026 |
| ChatGPTMulti-surface conversational AI product | ChatGPT web · ChatGPT apps | Instruction map; no full prompt | Aug 30, 2026 |
| CursorAI code editor, coding-agent platform, and early-beta git forge | Cursor editor · Cursor CLI · Cursor Cloud Agents · Cursor Origin | Instruction map; no full prompt | Sep 05, 2026 |
Publication posture is set for each page. Useful instruction intelligence may remain noindex when no complete first-party prompt artifact is established.
Evidence and provenance
How Promptcred handles evidence
Every artifact distinguishes the product from the model, preserves its source context, and states what the evidence does not establish. A public page can be useful without claiming to reproduce a complete prompt.
Read the methodology