What is an llms.txt file for a law firm?
It is a plain-text page at /llms.txt that names the firm and lists public pages. It is a map of the site. It is not a filing with an AI product, and it is not a payment.
By The FirmCited team, BizLegal AI Last updated
General information and software only - not legal advice, no attorney-client relationship. Consult a licensed attorney.
Reports are generated with AI-assisted software and may contain errors. Review before relying on them.
An llms.txt file is a plain-text page at the root of a website, usually at /llms.txt. It says what the site is and which pages a reader, including an automated reader, can open. For a law firm it is a map of the public site. It is not a request that an AI product name the firm, and publishing it does not produce that naming.
The filename is a community convention, alongside robots.txt. A person can open it in a browser. There is no login and no feed, and the file is meant to stay short. The URL is public, so write every line for a stranger and a crawler. A usable shape is a title, a short description, and a list of absolute URLs, so a reader can see the firm name, the place, and which pages to open.
Name the firm, the cities the public site covers, and the practice areas those pages discuss. List absolute https URLs with one line each. The home page, the practice pages, a short-answer page, and a contact line that matches the contact page are enough. Say that the pages are general information and that reading them does not create an attorney-client relationship. Do not claim a mention in an AI answer or a volume of new matters. Add a line when you add a page, and remove the line when you remove the page. A link that returns an error tells the reader the map is stale.
Publishing llms.txt does not insert the firm into ChatGPT, Perplexity, Claude or Google AI Overviews. Those products choose sources on their own. The file is not a fee and not a form you submit to a vendor. This page does not cite a Google ranking document that lists llms.txt as a ranking factor. A pitch that says the file places a firm inside an answer is a claim without a source. The file does not control crawlers: robots.txt does that. A sitemap is the longer URL inventory used for indexing. A firm can publish all three, because they do different jobs. The file also does not review a person's case. A model can still misstate the city or the practice, so compare a generated summary with the file.
The block below is an example for the fictional firm Northwind Sample Law. It is not a live file, and the links use example.com so they are not a real practice's pages. Copy the shape, then replace every line with your own firm, cities, and URLs. It has a name, a city, a practice, a limit on what the pages are, and four URLs. It does not ask an engine to cite the firm.
# Northwind Sample Law
> Estate planning practice in Portland, Oregon. This file lists public pages on this site.
Northwind Sample Law publishes general information about its estate-planning practice in Portland, Oregon.
Reading these pages does not create an attorney-client relationship.
## Pages
- https://example.com/ : Practice overview
- https://example.com/estate-planning : What the estate planning pages cover
- https://example.com/answers : Short answers to common questions
- https://example.com/contact : How to reach the office
## Contact
- Web: https://example.com/contact
Save the text as llms.txt in the site root so the public URL ends in /llms.txt. Serve it as text a browser can show, use https links, and open the URL in a private window to confirm a logged-out reader sees the full file. Click each link. Leave out client names, case facts, internal fees, and any sentence that promises a mention, a rank, or a volume of work. If a vendor offers to tune the file for citations, the file can change the map of the site. It cannot change what an answer engine writes. Tell the people who edit the site, so a redesign does not break the list unnoticed.
A firm can publish llms.txt and still be unnamed in an AI answer, and a firm can be named with no llms.txt file. The file answers "what does this site say it contains?" A mention check answers "did this engine name the firm, on this question, on this date?" Do not read one as proof of the other. The free AI Citation Scorecard is a 20-point self-check you score yourself. FirmCited does not run prompts on that page. Monitor repeats a 25-prompt list each month on up to 4 engines, sends a report when at least 2 engines answer, and emails a private portal link. The pricing page lists the fee in US dollars. This site's llms.txt describes FirmCited, software from BizLegal AI. FirmCited is not a law firm. The Northwind block is the example for a fictional law firm.
Open /llms.txt on the firm domain. Confirm the name, the cities, and each URL, and confirm the file does not offer a result the firm cannot produce. If a practice area was dropped, take it out in the same change. Reread the file when the site changes, as you would reread the footer. Serve one current file. When the file and the home page disagree, update the stale one.
The live file for FirmCited. It describes software, not a law firm.
A free 20-point self-check you score yourself. It does not run prompts.
A monthly 25-prompt list on up to 4 engines, with a private portal link by email, when at least 2 engines answer.
Current fees, in US dollars, for the audit, Monitor and the other plans.
How to record whether AI answers mention a firm, and what that record leaves out.
It is a plain-text page at /llms.txt that names the firm and lists public pages. It is a map of the site. It is not a filing with an AI product, and it is not a payment.
No. ChatGPT, Perplexity, Claude and Google AI Overviews choose sources on their own. The file can describe the site. It cannot require a mention, and it does not set a rank.
robots.txt sets crawl access. A sitemap lists URLs for indexing. llms.txt is a short reading list with a description of the site. A firm can publish all three. None of them replaces the others.
Leave out client names, case facts, non-public fees, and any line that promises a mention, a rank, or a volume of work. The file is public. Match what the public site already says.
This site publishes one at /llms.txt. It describes FirmCited, software from BizLegal AI, not a law firm. The example on this page is fictional and uses example.com.
25 prompts across up to 4 AI engines, competitor share of voice and a prioritised fix list, for USD 490, credited toward any install.