Prompting technique

Implicit Criteria Prompting

EvalAgent: Discovering Implicit Evaluation Criteria from the Web, arXiv:2504.15219 (2025) · read the paper

What it does

A two-step prompt improvement technique: first an LLM or search engine finds expert quality criteria for your task from authoritative sources, then these criteria are explicitly included in the main prompt as a technical specification. This transforms a vague request into a precise specification with professional quality standards.

Task typemulti step
Shapeiterative
Efforttwo or three steps
Addressesanswers that are close but wrong; the same prompt giving different answers
Published2025-04-21
AuthorsManya Wadhwa, Zayne Sprague, Chaitanya Malaviya et al.

When it helps

Use this technique when you need expert-level results from an LLM but you yourself are not an expert in the field and don't know all the quality criteria. Particularly effective for creative and professional tasks: scriptwriting, content creation, document preparation, strategy development. The method helps overcome the problem of 'sterile' and overly generic LLM responses by adding specific expert quality criteria found in authoritative sources. Ideal for situations

A prompt you can paste

Generic by design: it applies the technique without knowing your task. Adapt the marked parts.

Prompt
# Role
You are an expert prompt engineer specializing in applying the EvalAgent technique.

# Context
The user wants to understand and apply the EvalAgent technique, which focuses on discovering implicit evaluation criteria from expert sources to improve LLM responses. The user needs a reusable prompt template to guide this process.

# Task: Apply EvalAgent - Discovering Implicit Evaluation Criteria

You will guide the user through a two-step process to create a high-quality prompt using the EvalAgent methodology.

## Step 1: Research Implicit Criteria

**Objective:** To identify expert-level, "hidden" criteria that define a high-quality outcome for a specific task.

**Action:**
Formulate a "research prompt" to discover these criteria. Use the following template, replacing `[Your Specific Task]` with the actual task you want to accomplish.

**Research Prompt Template:**

If this one does not fit, the two closest alternatives in the corpus are Requirement-Oriented Prompt Engineering (ROPE) and Rebuttal Prompting, 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.

Worked example
# Role
You are an expert copywriter specializing in e-commerce product pages, with a knack for making technical descriptions sound natural and human.

# Context
I need to rewrite the product page description for a specific item. The current description is too technical and robotic.

# Task
Rewrite the provided product description to sound like a person wrote it, focusing on clarity, benefit-driven language, and an engaging tone.

# Product Description (Original - Technical)
[Please insert the original technical product description here]

# Implicit Criteria for "Human-Written" Product Page
Based on expert recommendations for effective product copywriting, the rewritten text must adhere to the following criteria:

1.  **Engaging Headline:** Craft a headline that immediately captures attention and hints at the core benefit or unique selling proposition (USP). Avoid jargon.
2.  **Benefit-Oriented Language:** Translate technical features into tangible benefits for the user. Instead of "X specs," explain "how X helps you achieve Y."
3.  **Relatable Tone:** Use conversational language, contractions (where appropriate), and a tone that resonates with the target audience. Avoid overly formal or robotic phrasing.
4.  **Storytelling/Scenario:** Briefly paint a picture of how the product fits into the user's life or solves a specific problem they might face.
5.  **Sensory Details (if applicable):** Incorporate words that appeal to the senses (sight, sound, touch, taste, smell) to make the product more vivid.
6.  **Clear Value Proposition:** Ensure the unique value or main advantage of the product is crystal clear and easy to understand.
7.  **Concise and Scannable:** Break down information into short paragraphs, bullet points, or easily digestible sentences. Use formatting to improve readability.
8.  **Implicit Social Proof/Trust Signal:** Subtly imply trustworthiness or customer satisfaction (e.g., "designed for ease," "trusted by many," without making unsubstantiated claims).
9.  **Clear Call to Action (Implied):** While not explicitly stating "Buy Now," the text should naturally lead the reader towards wanting to learn more or make a purchase.

# Output Format
Present the rewritten product description.

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.

Free · no signup · ~10s
0.00match confidence
single retrieval pass
Prompt for your task

      

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.

6,235techniques in the corpus
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 pageone technique, generic prompt
What you just ranone technique matched to your wording, nothing verified
Full runten 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

Related techniques

Requirement-Oriented Prompt Engineering (ROPE)ROPE (Requirement-Oriented Prompt Engineering) is an approach to creating prompts as detailed technical specif…Rebuttal PromptingRebuttal Prompting — a technique for correcting LLM errors through direct indication of inaccuracy. A simple r…MetaPrompting (Structured Reasoning)Method of prompting structuring by providing LLM with a ready-made response template in JSON or Markdown forma…Debugging Prompts FrameworkA systematic framework for diagnosing and preventing typical errors of language models in instruction understa…

All techniques · Failure modes and fixes