CLEAR
The only common framework with a slot for the thing most prompts are missing.
What is CLEAR?
CLEAR has five slots: Clarity, Language, Examples, Audience and Requirements. It is built for explanatory writing — documentation, guides, help-centre articles, anything whose job is to make a reader able to do something they could not do before.
One slot makes CLEAR worth knowing, and it is the E. Across every framework on this site, examples is the dimension prompts most reliably skip — CO-STAR has no slot for it, RACE has none, AIDA has none. It is worth ten points on the scorer and, more importantly, one short sample communicates format and register faster than any amount of description. CLEAR is the only common framework that puts a field there and makes you fill it.
Where CLEAR Came From
A modern acronym with no single author
CLEAR has no documented inventor and no history before prompting. It belongs to the family of acronyms assembled once people had catalogued which omissions actually degrade output — and unlike most of that family, whoever built it noticed that examples belonged on the list. Pages naming a founder are inventing one.
Note the name collision
There is an unrelated CLEAR coaching model — Contract, Listen, Explore, Action, Review — used in professional coaching and developed well before AI. If a search result is discussing coaching conversations, that is a different CLEAR.
What it is really for
CLEAR sits where technical writing sits: the reader has a job to do, is skimming, and will leave the moment they cannot find what they need. Its slots map onto the questions a documentation editor asks — is the task clear, is the language right for this reader, is there a worked example, who is this for, what must it contain.
The 5 Slots, One at a Time
Each slot is a decision. Leave it out and the model still makes it — just without you.
Clarity is Task with a bar attached. The strongest formulation names what the reader should be able to do afterwards — "so a developer can send their first successful request without asking us anything" is a testable condition, and it silently rules out the vague documentation that technically covers everything.
Language is where you set the level of explanation, and the useful move is to state assumed knowledge in both directions: what they already know, and what they do not. "Competent with HTTP, no knowledge of our product" tells the model exactly which terms need expanding and which would be patronising.
The slot that justifies the framework, and the one people still fill wrongly — describing an example is not giving one. Paste the real thing in a fenced block. On the scorer this is the difference between six points and ten, and in practice it is the difference between a model guessing your format and matching it.
Documentation has a second audience question most frameworks miss: not just who, but how. A reader who skims needs headings and scannable structure; a reader who works through sequentially needs continuity. Saying "will skim rather than read" changes the shape of the output more than naming the job title does.
Requirements carries the mechanical constraints — and in documentation the important ones are usually about completeness rather than style. "Every field must state its type and whether it is required" is the kind of rule that turns adequate documentation into usable documentation.
One Task, Before and After
The task: Document a new API endpoint. Both prompts below are scored by our free prompt checker — paste either in and you will get the same number, because the scoring is deterministic.
Write documentation for our new API endpoint.
Seven words. You will get plausible documentation for an endpoint that does not exist — invented field names, invented rate limits, invented status codes, all formatted immaculately.
Clarity: Document the POST /v2/webhooks endpoint so a developer can send their first successful request without asking us anything.
Language: Plain and direct. Expand every acronym on first use. Assume competence with HTTP and no knowledge of our product. Do not use the words simply, just or easily.
Examples: Include one complete request and two responses - a 201 and a 422. For example, the request block should look like this:
```
POST /v2/webhooks
Authorization: Bearer <token>
{"url": "https://example.com/hook", "events": ["invoice.paid"]}
```
Audience: Backend developers integrating us for the first time, who have not read any of our other documentation and will skim rather than read.
Requirements: 500-700 words in five headed sections - What it does, Authentication, Request, Responses, Common errors. Work through the failure cases before you write the happy path, so the error section is not an afterthought. Every field must state its type and whether it is required. Do not invent field names, rate limits or status codes that were not given to you.
Ninety-two, and the Examples slot is why. Replace the fenced block with a description of it and change nothing else: 82. Ten points for four lines.
Ten points for four lines
The gap between this prompt and an otherwise identical one is worth measuring, because it is the clearest demonstration on any of these pages of what an example is actually worth:
A cue phrase ("for example") plus a real delimited sample. Full marks on the Examples dimension.
"Include some examples of requests and responses" — same intent, no sample, everything else identical. Ten points gone, and the model now guesses your header format.
CLEAR has no persona slot. Correct for documentation, where house style should come from Language rather than from a character.
The lesson generalises beyond CLEAR: a cue phrase and a fenced block together are worth ten points, and most prompts give neither. If you use a framework without an Examples slot — CO-STAR, RACE, AIDA — add one anyway. It is the single cheapest improvement available to almost any prompt you will write.
Copy-Paste Prompt Template
Replace the bracketed placeholders with your specific details.
Clarity: [The task, and what the reader should be able to DO afterwards] Language: [Register, and assumed knowledge in both directions — what they know, what needs expanding] Examples: [An actual sample in a fenced block. Not a description of one] ``` [paste the real thing here] ``` Audience: [Who reads it, and HOW — skimming or sequential] Requirements: [Length, named sections, and one completeness rule. What must not be invented]
When CLEAR Fits — and When It Does Not
- API and technical documentation, where a worked example is not optional.
- Help-centre articles and troubleshooting guides.
- Onboarding material and internal how-to guides.
- Explaining something complicated to a reader who is not an expert.
- Any output where matching an existing format matters — the Examples slot is how you pin it.
- Persuasion. CLEAR has no slot for a reader’s motivation. Use AIDA or PAS.
- Making a case to a decision-maker. Use SCQA — CLEAR explains, it does not argue.
- Creative writing, where a supplied example constrains more than it guides.
- Very short output. Five slots for a two-line answer is more brief than deliverable.
10 Ready-Made CLEAR Prompts
Every prompt below was produced by the Frompting generator with CLEAR selected — not written by hand for this page. Each is scored by our prompt checker; the median is 90/100. Click one to open it, then copy.
API endpoint documentation 80
You are a technical writer tasked with creating comprehensive documentation for a new API endpoint. First, ensure the description of the endpoint’s purpose is precise and unambiguous, covering what the endpoint does and the problem it solves. Second, use concise, developer‑friendly language; avoid unnecessary jargon while employing correct technical terms, and define any term that may be unfamiliar on first use. Third, include concrete examples that illustrate a typical request and the corresponding response, showing all required fields, headers, and possible status codes. Fourth, write the documentation for the intended readers: [AUDIENCE: specify the primary developer audience, e.g., front‑end engineers, mobile developers, or integration partners]. Tailor tone and detail level to match their expertise. Fifth, meet the following requirements: the document must be formatted in Markdown, include a summary, endpoint URL, HTTP method, request parameters (including types and required/optional flags), authentication method, response schema, error handling, and at least one full request/response example. Limit the entire document to 400–500 words. If any of the above details are unknown—such as the endpoint’s exact path, required parameters, or authentication scheme—state your assumptions clearly and ask up to three clarifying questions before proceeding.
Refund policy in plain English 88
You are a documentation specialist tasked with drafting a concise, plain‑English explanation of a refund policy for a help‑centre article. The explanation should be written for [AUDIENCE: specify the primary users of the help centre, e.g., “customers who have purchased a subscription”]. Keep the total length between 150 and 250 words. Structure the content as follows: 1. Begin with a brief, one‑sentence statement that clearly defines what the refund policy covers. 2. Use simple, everyday language; avoid legal jargon and define any necessary terms in plain terms. 3. Provide one concrete example that illustrates how the policy works in a typical scenario. 4. Conclude with a short note directing readers to where they can find more detailed information or contact support. Quality criteria: - The text must be easy to read (aim for a Flesch‑Kincaid grade level of 8 or lower). - All statements should be factually accurate; if any policy detail is unknown, note it as “[UNKNOWN: insert specific policy detail]”. - The tone should be friendly and reassuring, encouraging trust. Boundary: Do not include any promotional language, pricing details, or legal disclaimer beyond what is necessary for clarity. If any essential details of the refund policy are missing, state your assumptions explicitly and ask up to three clarifying questions before finalizing the explanation.
Onboarding instructions for a new hire 96
You are a documentation specialist tasked with creating concise onboarding instructions for a new team member. Write clear, step‑by‑step guidance that a newcomer can follow without prior knowledge of your organization. Use straightforward language, avoid unnecessary jargon, and define any technical terms the first time they appear. Include brief illustrative examples for each major step to show exactly how the action should be performed. Tailor the tone and detail level to the intended reader, who is a [ROLE OF NEW MEMBER: e.g., junior developer, marketing associate, project coordinator]. The instructions should cover all essential elements needed for a smooth start, such as access setup, key contacts, primary tools, and initial tasks. Limit the entire document to 300–350 words, presented as a numbered list with optional sub‑points for clarity. Quality criteria: 1. Each step is unambiguous and can be executed independently. 2. Language is plain and any required terminology is briefly explained. 3. Examples directly illustrate the step they accompany. Exclude any company‑specific branding statements or promotional language; focus solely on functional onboarding content. If any critical detail (e.g., specific software name, internal system URL, or reporting line) is unknown, indicate it with a placeholder in the form [PLACEHOLDER: description of needed information]. Before finalizing, state any assumptions you are making and ask up to three clarifying questions about missing details.
Explaining a VPN to a non-technical reader 88
You are a clear‑thinking technical writer tasked with explaining what a VPN is to a non‑technical reader. First, define the concept in simple terms, avoiding jargon and focusing on the core purpose of a VPN. Second, choose language that is easy to understand: use everyday words, short sentences, and analogies that relate to common experiences. Third, include at least two concrete examples that illustrate how a VPN works in real life, such as protecting privacy on public Wi‑Fi and accessing content from another country. Fourth, tailor the explanation for an audience with no technical background, assuming no prior knowledge of networking or security. Fifth, ensure the response meets these requirements: - Length: approximately 200–250 words. - Structure: a brief introductory paragraph, followed by the examples presented as separate short paragraphs, and a concise concluding sentence. - Quality: clarity of the main idea, simplicity of language, relevance of examples, and suitability for a non‑technical reader. Do not include technical specifications, code, or advanced networking terms. If any assumptions are needed, state them explicitly.
How-to: setting up two-factor authentication 91
You are a technical documentation specialist tasked with creating a concise, step‑by‑step how‑to guide for setting up two‑factor authentication (2FA). The guide should be written in plain, direct language that avoids unnecessary jargon while clearly defining any technical terms the first time they appear. Include concrete, real‑world examples illustrating each major step (e.g., enabling 2FA via a mobile authenticator app, configuring SMS codes). Target the guide to [AUDIENCE: specify the primary readers, such as non‑technical end users, IT support staff, or security‑aware managers]. The final output must be a single markdown document of approximately 350 words, organized with numbered steps, brief explanatory paragraphs, and example snippets where appropriate. Quality criteria: 1. Every step is actionable and ordered logically. 2. Language is clear, with any required technical terms defined inline. 3. Examples directly support the corresponding steps and are easy to follow. Exclude any platform‑specific branding or references not provided by the user. If any of the above details are unclear, state your assumptions and ask up to three clarifying questions before drafting the guide.
Release documentation for a software update 92
You are a technical writer tasked with creating comprehensive release documentation for a software update. The documentation must be concise, use precise terminology, and include concrete examples that illustrate each new feature or fix. It should be tailored to the intended readers and satisfy all necessary informational requirements. [PRODUCT NAME]: provide the name of the software product. [VERSION NUMBER]: specify the version being released. [KEY CHANGES]: list the major new features, improvements, and bug fixes. [TARGET AUDIENCE]: identify who will read this document (e.g., end‑users, administrators, developers). Produce the documentation in markdown with the following structure: 1. A brief introductory paragraph that clearly states the purpose of the release. 2. A bulleted list of new features, each followed by a short, concrete example of its use. 3. A bulleted list of improvements, each with an illustrative scenario. 4. A bulleted list of bug fixes, each describing the issue resolved and its impact. 5. A “Getting Started” section with step‑by‑step instructions for installing or upgrading to the new version. 6. A “Known Issues” section, if any, summarizing remaining limitations. The entire document should be between 300 and 500 words. Ensure the language is straightforward, avoids unnecessary jargon, and defines any technical terms on first use. Do not include any information beyond what is provided; if any of the placeholders above are missing, state the assumption you are making and ask up to three clarifying questions before finalizing the document.
Explaining an insurance term to a customer 68
You are an insurance communications specialist. Your task is to explain a specific, complex insurance term so that a customer can fully understand it. First, restate the term in plain language, removing any jargon. Second, choose wording that is simple, conversational, and avoids technical jargon while still being accurate. Third, provide one or two concrete, relatable examples that illustrate how the term applies in real‑world situations. Fourth, tailor the explanation to the customer’s background and needs. Deliver the explanation in a single paragraph of approximately [WORD COUNT: e.g., 120‑150] words, written in a friendly tone appropriate for a [CUSTOMER TYPE: e.g., non‑technical adult, small‑business owner, senior citizen]. Ensure the content is clear, jargon‑free, and includes the examples, and explicitly state any assumptions you are making about the term or the customer. If any key details are missing, ask up to three brief clarifying questions before completing the explanation.
Troubleshooting: printer will not connect 90
You are a technical writer tasked with creating a concise troubleshooting guide for a printer that will not connect. Write the guide in clear, step‑by‑step prose suitable for a user who needs to resolve connectivity issues quickly. Use plain language, avoid unnecessary jargon, and include brief illustrative examples for each major step (e.g., “If the Wi‑Fi indicator blinks, try restarting the router”). The guide is intended for [AUDIENCE: specify the typical user, such as office staff, home users, or IT support personnel]. If the printer model or connection type is relevant, insert placeholders like [PRINTER_MODEL] and [CONNECTIVITY_TYPE] where appropriate. Deliver the guide as a single markdown document no longer than 350 words. Structure it with a brief introductory sentence, followed by a numbered list of troubleshooting steps, each step limited to two sentences and accompanied by a short example in parentheses. Ensure the guide: 1. Is easy to scan and follow. 2. Uses consistent terminology for hardware components. 3. Provides at least one concrete example per major step. Exclude any brand‑specific marketing language or warranty statements. If any assumptions are needed, state them explicitly at the end of the guide. Before writing the final answer, work through the problem step by step and weigh the main trade-offs; present only the reasoned conclusion, not your working notes.
Instructions for submitting an expense claim 100
You are an expert technical writer tasked with creating a concise, step‑by‑step guide that explains how to submit an expense claim. The guide should be written in plain, direct language, avoiding unnecessary jargon while defining any required terms. Include a short, concrete example that walks through a typical claim from receipt collection to final approval. Tailor the tone and detail level for the employees who will use the guide, assuming they have basic familiarity with company tools but need clear procedural instructions. Ensure the guide covers all mandatory elements such as required documentation, approval workflow, submission deadlines, and any system‑specific fields. Output: a single markdown document of approximately 300 words, organized with numbered steps, a brief example block, and a final checklist. Quality criteria: 1. Each step is unambiguous and actionable. 2. Technical terms are defined at first use. 3. The example accurately reflects the process and reinforces the steps. Exclude any company‑specific branding, policy details, or system URLs that are not provided. If any of the following details are unknown, insert a placeholder in the format **[PLACEHOLDER: description]** and proceed with the assumption that the missing information will be supplied later: - [SYSTEM_NAME: name of the expense submission platform] - [APPROVER_ROLE: title of the person who approves claims] - [DEADLINE: typical submission deadline (e.g., 30 days after expense)] - [REQUIRED_DOCS: list of documents required (e.g., receipts, mileage log)] State any assumptions you make and ask up to three clarifying questions before finalizing the guide.
Explaining compound interest to a teenager 83
You are a friendly financial educator. Explain the concept of compound interest in a way that a teenager can easily understand. Begin with a concise definition that avoids jargon, then use simple, everyday language to describe how interest builds over time. Follow with at least two relatable examples—one using a savings account and another using a common teen scenario such as a video‑game purchase or a part‑time job earnings. Tailor the tone and style to a typical teenager aged [AGE: specify the teen’s age range, e.g., 13‑18] who may have limited prior knowledge of finance. Keep sentences short, vary length for natural flow, and define any necessary terms the first time they appear. The explanation should be approximately 250–300 words, presented as a single cohesive paragraph followed by a bullet‑point list of the examples. Ensure the content is clear, engaging, and free of unnecessary technical detail. Do not include any promotional language, references to specific banks or products, or advanced mathematical formulas beyond the basic compound interest formula. If any assumptions are needed, state them explicitly. If you need clarification on the teen’s background or preferred example topics, ask up to three brief questions before completing the response. Write this for [AUDIENCE: who will read the output, and how much they already know]. Match the depth, vocabulary and examples to that reader.
Scores range from 68 to 100. They are shown as generated rather than cherry-picked — a library where every entry scores in the nineties tells you it was curated, not measured.
CLEAR vs the Alternatives
The better choice when voice matters — CO-STAR has Style and Tone. CLEAR trades those for the Examples slot, which is usually the right trade for documentation and the wrong one for a newsletter.
The closest relative, and also carrying an Examples slot. TRACE swaps CLEAR’s Language for Context — pick by whether the register or the background is the harder part.
Better when writing as a specific person for a specific team. CLEAR is reader-first, RACE is writer-first.
For transforming supplied material rather than explaining a subject. ICIO has a dedicated Input slot; CLEAR assumes the knowledge is the model’s.
Not a rival — the technique CLEAR’s E slot formalises. If you find yourself needing three or four examples rather than one, you are doing few-shot prompting and should lean into it.
Five Ways People Get CLEAR Wrong
The defining CLEAR error, and it costs the framework its main advantage. "Include some examples" is a description; a fenced block is an example. Measured: ten points.
"Professional and easy to understand" sets nothing. Language works when it states assumed knowledge in both directions — what they know, and what needs expanding.
Skimmers and sequential readers need different documents. It is the highest-value half of the Audience slot and the half that gets left out.
"Comprehensive and well organised" is not a requirement. Named sections, a word range, and one completeness rule such as "every field states its type".
Documentation prompts invite fabricated field names, rate limits and status codes, and the output looks authoritative either way. Forbid it explicitly.
CLEAR Questions
What does CLEAR stand for?
Clarity, Language, Examples, Audience, Requirements. Five slots aimed at explanatory writing.
Who created the CLEAR framework?
No documented author. It is a modern prompt-engineering acronym. Note the collision with the unrelated CLEAR coaching model — Contract, Listen, Explore, Action, Review — which predates AI and appears in the same search results.
Why does CLEAR score higher than CO-STAR on the same kind of task?
The Examples slot. Examples is worth ten points and CO-STAR has nowhere to put one, so a well-filled CLEAR prompt starts ten points ahead before anything else is written.
What counts as a real example?
Something a reader could copy. A fenced block, a quoted sentence, a sample row. A description of what the example would contain does not count, and the difference is measurable.
Should I use CLEAR for marketing copy?
No. It has no slot for tone or for the reader’s motivation, and both matter more than anything CLEAR offers. Use CO-STAR, AIDA or PAS.
Can I add an Examples slot to other frameworks?
Yes, and you should. Nothing stops you appending a fenced sample to a CO-STAR or RACE prompt, and it is the cheapest improvement available to either.
Generate a CLEAR Prompt Instantly
Skip the manual template — Frompting applies CLEAR to your topic in one click.
Try it FreeFramework Details
| Name | CLEAR |
| Stands for | Clarity-Language-Examples-Audience-Requirements |
| Domain | Communication & Storytelling |
| Steps | 5 |
| Access | Free |