Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesTo turn website content into an llms.txt file, choose the pages that best explain your site, write a concise Markdown guide with useful links to them, and publish it at the root of your site or in the path it describes. Treat it as a curated guide for AI systems—not a dump of your website, a replacement for robots.txt or sitemap.xml, or a guarantee that any system will read or cite your pages.
What an llms.txt file does
llms.txt is a proposal for a Markdown file that gives language-model-based agents context about a website and directs them to relevant material. Its author, Jeremy Howard, describes it as a way to provide LLM-friendly content; the proposal is published at llmstxt.org. The idea is to offer a compact orientation to a site whose useful information might otherwise be spread across pages with navigation, advertising, or JavaScript.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
The Handbook of Technical Writing with 2020 APA Update | $60.69 | Buy on Amazon |
| 2 |
|
Handbook of Technical Writing, Tenth Edition | $38.46 | Buy on Amazon |
| 3 |
|
The Handbook of Technical Writing | $44.99 | Buy on Amazon |
| 4 |
|
The Technical Writer's Handbook: Writing with Style and Clarity | $41.98 | Buy on Amazon |
| 5 |
|
The Insider's Guide to Technical Writing | $35.95 | Buy on Amazon |
The file is a guide, not a full representation of your site. It can explain what a project is, point to key documentation, and tell a reader how sections fit together. The linked pages remain the detailed sources. A useful file therefore makes deliberate choices about relevance and context rather than attempting to include every URL or copy every page.
How it differs from robots.txt and sitemap.xml
sitemap.xmlis an index of pages intended for discovery;llms.txtis a selective guide to content worth consulting.robots.txtcommunicates crawler access preferences. Anllms.txtfile does not grant access, deny access, or replace robots rules.- The format is a proposal or convention, not a ratified web standard. Publishing a file does not require an AI system to fetch it, use it, or cite your site.
Choose the scope before you write
Decide whether the guide should describe the whole website or only one part of it. A site-wide guide normally belongs at /llms.txt. A guide for a particular section can sit within that path, such as an llms.txt file in a documentation directory. The proposal allows either approach: a file describes pages under its own path, and the more specific applicable file should be used for that scope.
Recommended Free Tools
#1 Best Overall
Keep the scope coherent. If your site combines a product, documentation, a support center, and a blog, a single root guide can point to the most important resources in each area. If a section has its own audience and substantial documentation, a scoped guide can orient agents specifically to that material. Avoid creating multiple guides whose coverage is unclear or whose links compete without explanation.
Curate the pages agents should consult
Make a short list of pages that explain the subject better than a generic crawl would. Depending on the site, these may include project context, getting-started material, API documentation, policies, or a small number of important background pages. For each candidate, ask whether a reader needs it to understand the site and whether its link and description will still be accurate as the site changes.
- Choose pages for usefulness, not merely because they appear in the sitemap or have high traffic.
- Prefer an authoritative page over several overlapping pages on the same subject.
- Use descriptive link text and a brief note that tells a reader what the destination contains.
- Remove stale, duplicate, or irrelevant destinations rather than exporting them automatically.
- Keep the guide brief enough to navigate; link to detailed material instead of copying it all into the file.
A sitemap can help you discover candidate URLs, but it is not a finished llms.txt. A sitemap aims to provide an index; this file should explain which resources matter and why.
Write the file in Markdown
Start with an H1 naming the site or project. The proposal specifies that heading as the only required section. Add a concise blockquote with context, then plain Markdown guidance if the structure of the site needs explaining. Use H2 headings to group links by subject. Under each heading, add Markdown links with short, accurate descriptions. An Optional section can separate secondary resources from core material.
Rank #2
This illustrative outline uses example URLs rather than pages from a particular real site; replace them with destinations you have verified:
# Site or project name
> One or two sentences giving an agent the context it needs.
Brief guidance on the site's structure and how to interpret its pages.
## Core documentation
- [Getting started](https://example.com/docs/start.md): Setup and first steps.
- [API reference](https://example.com/docs/api.md): Detailed interface documentation.
## Policies
- [Privacy](https://example.com/privacy.md): Privacy practices.
## Optional
- [Background article](https://example.com/blog/background.md): Additional context.What belongs in the context paragraph
State what the site or project is and, if needed, how its material is organized. Do not use this space to repeat the link list or make claims that the linked pages do not substantiate. If an agent needs to know that a documentation section covers a particular product or audience, say so plainly and link to the relevant authoritative pages.
How to group and describe links
Choose headings that reflect how a reader would look for information: for example, core documentation and policies, as in the outline above. Link names should identify the destination, and a note after the link should add useful context rather than restating the title. Keep optional or supplementary reading distinct so that the core references remain easy to identify.
Link to clean Markdown when available
The proposal recommends linking to clean Markdown versions of pages where a site provides them. A site may expose a Markdown copy using a path convention such as an appended .md or an extension-replaced .md path; verify the actual address rather than assuming either convention works. For directory URLs, the proposal also describes using a Markdown file convention. The destination should resolve and contain the material its label promises.
Rank #3
Where a page has a Markdown alternative, the proposal describes identifying it with an HTML rel="alternate" link. It also describes rel="describedby" links from pages to the covering llms.txt. Those relationships can be expressed in HTML or HTTP Link headers. They are optional conventions in the proposal, not a requirement for the basic file. If you add them, make sure their targets match the Markdown page and the guide that actually covers the page.
Create the file manually or use a generator
Manual authoring is enough to create a valid, useful guide: choose the pages, write the Markdown, and publish it. A generator or platform plugin can help assemble a draft, but it is an optional convenience—not a prerequisite and not a substitute for editorial review. The proposal lists implementation resources including a WordPress plugin and a Docusaurus plugin; that does not establish that a particular tool is necessary, accurate for every site, currently maintained, or affiliated with the proposal.
| Approach | Useful when | What to review |
|---|---|---|
| Write it manually | You want direct control over a small, curated guide. | Check every destination, description, and scope yourself. |
| Start from a generator or plugin | You want an initial draft or a workflow connected to a site platform. | Review page relevance, stale or duplicate URLs, descriptions, path coverage, and Markdown destinations before publishing. |
A currently surfaced service describes generating a file from a sitemap, but the proposal and available information do not establish that any generator produces an accurate result for every site. Treat generated output as a draft. A comprehensive export can include low-value or duplicate pages and miss the context that makes a curated guide useful.
Publish and check the result
- Save plain Markdown text. Name the file
llms.txtand make sure it is served as a text file rather than rendered as a webpage or saved with an extra extension. - Place it at the scope you chose. Publish at the site root for a site-wide guide, or under the relevant path for a section-specific guide. Confirm that the file’s location matches the pages it describes.
- Open the published address. Visit the exact URL in a browser or request it from a client you use for checking site content. Confirm the response shows the intended Markdown rather than an error, redirect to an unrelated page, or HTML template.
- Check every link and note. Follow each destination, confirm it resolves to the intended page, and verify that the note accurately describes it. If you link to a Markdown alternative, confirm that alternative exists.
- Compare the guide with the site. Confirm it represents the selected pages and the intended path scope. Do not assume a sitemap export or plugin output has made these editorial checks for you.
- Keep it current. Revisit the file when important pages move, are renamed, or become obsolete. Update link text and descriptions when the destination’s purpose changes.
The proposal does not prescribe a hosting provider, CMS deployment method, or required validator. Use the publishing process that fits your site, then verify the public result directly.
Free tools Windows power users keep installed
One-click scans. No signup required.
Rank #4
- Used Book in Good Condition
Common problems and fixes
The file returns a not-found response
Check the exact published path, capitalization, and filename. A root guide and a section guide have different addresses; confirm that your deployment put the file where you intended rather than only adding it to a local project folder.
The URL displays a web page instead of the Markdown
Check whether the site or CMS is rewriting the path to a page template or redirecting it. Adjust the publishing configuration so the intended plain-text Markdown file is served at the chosen address, then open that address again to verify the actual response.
Links work for people but do not provide clean Markdown
A normal page link can still be useful, but it may contain the surrounding interface the proposal aims to avoid. If a clean Markdown version exists, link to the verified alternative. Do not guess a filename suffix or path convention: test the candidate destination first.
The file is long, repetitive, or reads like a sitemap
Remove low-priority and duplicate entries, group the remaining links by reader need, and put detail in the destination pages. The goal is orientation and selection, not a complete inventory.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Best Value
A plugin or generator includes irrelevant pages
Edit the generated draft. Remove pages that do not help explain the site, consolidate overlapping links, and check descriptions and scope. The presence of a URL in a sitemap does not make it a core reference.
Publishing the file does not change AI visibility
Do not treat publication as a ranking tactic or a guarantee of discovery. The proposal gives a format for a guide; it does not establish a measured traffic or citation benefit. Make the file accurate and useful independently of whether a particular system chooses to use it.
What an llms.txt file can and cannot promise
The proposal’s v2 page is dated September 3, 2024, and lists a modification date of August 10, 2026. Its background says that thousands of sites publish an llms.txt file; that is an adoption statement from the proposal, not an independently established count or evidence of resulting visibility. No percentage, traffic lift, or universal adoption effect should be inferred from it.
A well-maintained file can give an agent concise context and direct it to selected, relevant resources. It cannot guarantee that a crawler will fetch the file, that a model will use a linked page, or that the page will be cited. Maintain ordinary site navigation, a sitemap, and robots rules for their distinct purposes; use llms.txt as an additional curated guide.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Or skip the browser setup
ScreenshotNeo is a website screenshot API and MCP server, not an llms.txt generator: use the steps above to write and publish the guide. If you separately need a screenshot of a site page for visual review, one GET request can return an image or PDF. The API can accept consent banners and remove known consent platforms, newsletter popups, and chat widgets before capture. Bot checks, blank pages, timeouts, and failed loads are never billed, and cache hits cost nothing; responses include X-Page-Verdict and X-Billed headers. Its MCP server provides take_screenshot, get_page_info, and capture_pdf for AI agents.
curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://stripe.com -o shot.webpSee the ScreenshotNeo API documentation for request options. The free plan includes 1,000 screenshots per month with no card; paid plans start at $5 for 3,000. Sign up for ScreenshotNeo’s free plan.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
