Hard-to-Easy Instruction Ordering
What it does
A technique for ordering instructions in a prompt following the "hard-to-easy" principle. Complex constraints (word prohibitions, structure) are placed at the beginning of the instruction list, simple ones (capitalization, final phrase)—at the end. This forces the model to immediately account for difficult requirements during generation, improving accuracy of meeting all conditions.
When it helps
Use this technique when a prompt contains multiple instructions and constraints, especially if the model fails to execute them simultaneously. Ideal for tasks with complex requirements for text format, style and content. Apply when creating content with strict constraints (word prohibitions, character limits, specific structure). Especially effective for generating text where compliance with all conditions simultaneously is important. The method is universal and works with any LLM without
A prompt you can paste
Generic by design: it applies the technique without knowing your task. Adapt the marked parts.
**Role:** You are an AI researcher specializing in prompt engineering and LLM behavior.
**Task:** Generate a prompt that demonstrates the "Hard-to-Easy Instruction Ordering" technique. The prompt should guide an LLM to perform a complex task with multiple constraints, ordering these constraints from most difficult to easiest for the LLM to follow.
**Context:** The user wants to adapt the "Hard-to-Easy Instruction Ordering" technique for a general-purpose prompt. This prompt should serve as a template that a user can easily modify for their own tasks.
**<INSTRUCTIONS>**
Here are the rules for generating the prompt, ordered from most difficult to easiest for an LLM to follow:
1. **[Very Difficult] Constraint Prioritization & Nuance:** Define and strictly enforce the most complex constraints first. These should require deep understanding and self-correction throughout generation, such as nuanced tone requirements, specific stylistic exclusions, or complex logical interdependencies between elements. Explicitly state that these are the highest priority and must be adhered to from the outset.
2. **[Difficult] Content Generation with Specific Exclusions/Inclusions:** Generate the core content of the response, but with a high degree of specificity. This includes mandatory inclusion of certain elements or concepts, and strict exclusion of others (e.g., specific keywords, phrases, or topics). This requires careful content filtering during generation.
3. **[Medium] Structural Formatting & Length Control:** Apply formatting rules that affect the overall structure and length. This could include word count limits, specific section requirements (e.g., "must have an introduction and conclusion"), or the use of particular data structures (like tables or bullet points).
4. **[Easy] Simple Lexical/Stylistic Rules:** Implement simple, surface-level stylistic rules that are easy to apply once the core content is generated. Examples include using lowercase only, ending with a specific phrase, or starting with a predefined sentence.
5. **[Very Easy] Final Output Confirmation:** Conclude the response with a simple, predefined confirmation or concluding remark.
****
**Placeholder Instructions for User Adaptation:**
* **[Very Difficult] Constraint Prioritization & Nuance:**
* **[User Action]:** Replace this section with your most complex, nuanced constraints. Examples: "Maintain a tone of empathetic skepticism, avoiding any overly optimistic or dismissive language." or "Ensure that every point made directly addresses the core user need, and implicitly refutes common misconceptions about [topic]."
* **[User Action]:** Clearly state the priority: "These are the top priority constraints."
* **[Difficult] Content Generation with Specific Exclusions/Inclusions:**
* **[User Action]:** Define the core task here (e.g., "Generate a summary of the provided document," "Write a marketing email," "Explain a complex concept").
* **[User Action]:** Specify inclusions: "Include at least three actionable recommendations." or "Mention the historical context."
* **[User Action]:** Specify exclusions: "Do not use jargon related to [field]." or "Avoid mentioning competitors by name."
* **[Medium] Structural Formatting & Length Control:**
* **[User Action]:** Specify length: "The final output should not exceed [Number] words/paragraphs."
* **[User Action]:** Specify format: "Present the information as a numbered list." or "Use Markdown for headings and bullet points."
* **[Easy] Simple Lexical/Stylistic Rules:**
* **[User Action]:** Add simple stylistic rules: "Write entirely in lowercase." or "Ensure all sentences start with a verb."
* **[Very Easy] Final Output Confirmation:**
* **[User Action]:** Specify the exact concluding phrase: "End the response with the exact phrase: '[Your desired concluding phrase]'."
**Begin generating the prompt based on these placeholders.**
If this one does not fit, the two closest alternatives in the corpus are Structured Prompting with Iterative Refinement and Requirement-Oriented Prompt Engineering (ROPE), which target the same failure from a different angle.
Worked example
The same technique applied to a concrete job: rewrite a product page so it reads like a person wrote it. Use it as the pattern for your own case rather than as a finished artefact.
**Role:** You are an expert copywriter specializing in transforming technical product descriptions into engaging, human-readable content. **Context:** You need to rewrite a product page for a new product. The original content is likely technical and may not resonate with a general audience. The goal is to make it sound like a person wrote it, focusing on benefits and user experience rather than just features. **<INSTRUCTIONS>** Apply the "Hard-to-Easy" instruction ordering principle to ensure the highest quality rewrite. Arrange the following instructions in order of difficulty, from most challenging for an LLM to least challenging: 1. **[Very Difficult] Natural Language Emulation:** Rewrite the product description to sound as if a human, not an AI, wrote it. Focus on conveying enthusiasm, relatability, and a genuine understanding of the user's needs. Avoid jargon, overly complex sentence structures, and robotic phrasing. Use contractions and conversational tone where appropriate. 2. **[Difficult] Benefit-Oriented Transformation:** Shift the focus from technical features to user benefits. For every feature mentioned or implied, explain *why* it matters to the user and *what problem it solves* for them. 3. **[Medium] Clarity and Conciseness:** Ensure the final text is clear, easy to understand, and avoids ambiguity. Remove redundant information and streamline sentences. Aim for a length that is engaging but not overwhelming. 4. **[Easy] Tone Consistency:** Maintain a consistent, positive, and approachable tone throughout the entire rewritten text. 5. **[Very Easy] Final Polish:** Perform a final read-through to catch any grammatical errors or awkward phrasing that might have slipped through. **** Rewrite the following product description: [INSERT PRODUCT DESCRIPTION HERE] Begin the rewrite.
Get this written for your actual task
Paste what you are trying to do and the corpus will be matched against it directly. Free, no account, about ten seconds.
single retrieval pass
That number is low on purpose, and it is real. It is the raw similarity of one retrieval pass: no specialist read the paper, no judge compared anything, the first plausible match won.
one of which is this page
Picking the right one for a specific task is the work, and it is the work GetDecision does.
| This page | one technique, generic prompt |
| What you just ran | one technique matched to your wording, nothing verified |
| Full run | ten specialists read the papers in full, a judge ranks the top three for your task and shows its reasoning, generation on the model you pick, saved to your history |
See the top three for your taskTen specialists, a judge, and the reasoning shown. Free account, first run included.
Run the full analysis