Githubbing Law: Open-source legal doc repositories

What would a Github for Law look like? And is it worth building?

This afternoon I got an email from a site visitor who asked if I knew of any projects in the works that stakes a Github for Lawyers out — and if there is a profitable business model in such an undertaking.

Github is a site for coders to upload code they’re authoring & collaborating on, and that allows other members of the community to ‘fork’ from their code.  Meaning, others can take the work that one person has created & then make separate versions of it, stored separately but still linked back to the original.

The site also allows for collaborative editing. A member can request to change a document — calling out an issue, or proposing a better way to accomplish a task — by submitting a ‘pull request’. The author can review the request & approve it if she agrees.

Github can serve as an advanced filing system — keeping track of revisions, allowing for multiple, customized versions of a single type of document, and allowing for collaborative back & forth.  The site can be used not only for code files, but also standard word docs or any other file types.

The other thing about Github is an ethos of openness & transparency.  Any Internet user can see the work product, and anyone who signs up for an account can fork the work product off & customize it in a new version. The Github model aims for a collaborative work flow, that prioritizes opening up your work product to the community, letting others borrow it & edit it, and taking in criticisms from community-members who think they can build off your work to improve it.

So could the Github model be applied to legal professionals’ work, or how the legal services that non-lawyers use?

My short answer to the visitor: there are an increasing number of Legal Document Repositories, many of them now overlaid with a user-friendly interface that allows the user to take the standard document and fill in the designated fields with their own information.

Thus, the user can take a standard doc and make it her own by simply entering in a few pieces of information (that she likely has at her finger tips). Some of these document repositories are even hosted in part on Github, so that any other visitor who signs up for Github could fork these documents & customize them for her own.

Here is a short inventory of projects that are creating such open repositories + form-filling interfaces.

One noticeable thing: most of them are aimed at entrepreneurs as the main user.  The use case is someone setting up a start-up & trying to get right with corporate law.

(There is another branch of Legal Document repositories: government & legal service sites that compile standardized legal forms for users. These sites mainly just create a searchable/browsable list of downloads of government forms. I haven’t included them in this post, but the design of those sites deserve their own separate examination in a later post.)

Docracy is perhaps the most advanced repository online right now.  It allows for any member with an account to upload a legal document that will be shared with the rest of the community. Any other member can then make a private customization of the document, which they can then invite others to work on & sign. Other members can also fork the document publicly, to edit & customize the doc publicly, allowing other members to use their version.

Open Law Lab - Github for Law - Docracy 1 Open Law Lab - Github for Law - Docracy 2 Open Law Lab - Github for Law - Docracy 3 Open Law Lab - Github for Law - Docracy 5


Restatement is a project from Jason Boehmig, Tim Hwang, and Paul Sawaya, funded through the Knight Foundation’s Prototype Fund.  They are building a repository of legal documents along with an interface-overlay to fill in the docs’ fields in a user-friendly way. They also promise that the site will allow users to use their document repository with other tools — to be able to run analytics, to be able to auto-fill and auto-create documents, and otherwise ‘hack and slice’ a legal document. Right now Restatement is not fully live, but it does have some demos.
Open Law Lab - Github for Law - Restatement 1

Open Law Lab - Github for Law - restatement 2


A similar project is coming out of Singapore, from Cofounders Pte. Ltd., with its Legal Boilerplate collection of documents. Their site is more populated with legal documents, each of which the user can fill out in their browser by scrolling through the document & entering numbers and words into the open fields. The site also has a Github repository attached to it.

Open Law Lab - Github for Law - Cofounders Legal 2Open Law Lab - Github for Law - Cofounders Legal 3

Open Law Lab - Github for Law - Cofounders Legal 1


Law firms have also created start-up oriented legal document repositories.  Cooley has a set of open source legal documents available on the startup-accelerator Techstars’ website.

Open Law Lab - Github for Law - Cooley and Techstarts


Fenwick & West has the Series Seed repository available on its own project website, along with a Github page.

Open Law Lab - Github for Law - Series Seed 1

Open Law Lab - Github for Law - Series Seed 2

From this quick inventory, the status quo model is clear:

  1. collect together some standard legal documents
  2. identify the fields that any user will have to change to customize the document to their situation
  3. build an interface in the browser that lets the user enter their custom data into these fields
  4. allow the user to export the document for their own use
  5. perhaps, make the document open for forking & editing, for others to customize, build upon, or improve

The question is what the next-level ambitions for Legal Document Repositories should be.

One path would be scaling up the types of documents available (like Docracy is doing, with different types of crowdsourced docs besides startup-focused ones, though their site is still mostly populated with startup-docs).

Another would be what Restatement is hinting at, with more integration of the legal documents into automated processes, analytics, and other work flows that could prove to increase efficiency & knowledge for legal professionals.

I’d love to see some more features & functions that serve the end-user of these documents.  Can we have some metrics associated with these docs?  Docracy has a love button, that allows a visitor to see how many people ‘love’ this document, and it has stats on downloads and views as well.  But there must be a better metric than just ‘popularity’.

Could we have Expert Curators who give their reviews & recommendations of documents?

Could we have Stats on the Document’s Worth, that document the number of lawsuits or other problems that arise after using a document?

Could we have Popular Ratings, Yelp-style, in which those who have downloaded the site are then invited back a year or two later to give star-ratings & comments to the document they have been using?

Could we have Customized Forks of Documents with clear descriptions of which scenario this version should be used for?

Could we have Best Practice Packages of documents, that have been vetted and staged so that a user who is on a certain legal path can use this series of documents and know that expert users recommend them for exactly the scenario she’s in?

Just some ideas — I’m sure there are plenty more improvements for stage 2 of these legal document repositories.  Post away in the comment section, or send me an email!



Great post–it seems like all of these repository attempts are trying to target startups and other non-lawyers. They’re also geared toward transactional issues. Github is largely a community by coders and for coders. I think a good start for something like this would be in a similar vein–by lawyers, for other lawyers. I could see a site for litigators to post motions, pleadings, etc, that can then be forked by other lawyers for every conceivable type of lawsuit doing really well. Eventually, so many topics would get covered in so many different ways that even non-lawyers might find it useful in small claims and other pro se situations.


I think we took a shot at very close to what you are speculating about with, but ran out of runway (among other things) before we could work it all out. Docracy is GREAT as a github for legal docs, but it suffers from exactly the problems that github does when it comes to figuring out not only what is good code, but also good code for you, unless you are already a coder. Obviously, that is a problem if you are trying to provide legal docs for nonlawyers, and you already hit on most of the major problem areas. I think there is a tremendous opportunity out there, and someone will get it right. Casetext may be closest, but it doesn’t appear to be there primary focus at this stage.

Anyhow, if you care to discuss further, please feel free to drop me a line. We did a lot of research on this problem before we dove in.

Awesome post, Margaret. I think there’s huge potential in these ideas, though like Kyle I tend to favor approaches designed for the developers (the lawyers) rather than the end users (the clients). To take the analogy a step further, clients really need a useful, functional end product. In most situations, they can’t do much with a snippet or prototype or proof of concept, and they could do harm in some situations. Lawyers, on the other hand, can take a little piece of “code” from a brief or opinion or law journal or document repository and then adapt or improve it. In fact this is exactly what happens in law practice everyday. To me at least, the most exciting challenge is how to get people to think about legal work itself more in terms of open source and open innovation. And the benefits of that, to clients and others, could be enormous.

Would be interesting to see if such a repository could help transform the bloviating and overblown style and syntax of traditional contracts and law documents into something approaching an intelligible vernacular.

Hi Margaret,

My name is Julio Avalos, GitHub’s General Counsel/Chief Legal Officer.

Really enjoyed the post. I think that importing OSS workflows into legal practice is critical. GitHub Legal operates on GitHub, for instance, and tries as much as possible to work using machine readable plain text/Markdown, Pull Requests, and the like. I’m sure that millions (if not billions) of dollars are wasted every year just by merit of the fact that the industry is tied to MS Word and working essentially the same flow that was used by attorneys twenty or thirty years ago. Back in my law firm life, certainly entire days worth of work were lost when someone discovered that they’d been working in the wrong version of a brief, or because it was impossible to tell what the latest version of a document was, or what nomenclature we should be using when re-naming a file. In this day and age, it’s an embarrassment.

There’s every reason to believe that automation, OSS version control, and open sourcing legal documents will lead to higher quality legal work product, the breaking up of law firm cartels, a drop in the cost of legal services, and wider access to legal representation for marginalized, disenfranchised, or just plain average citizens. Currently, the graduate of a “lower tier” law school or really anyone that isn’t willing or able to work for one of the posh law firms has no ability to get in front of the in-house counsel of any major tech (or other) company. As in a true monopoly, companies are forced to go through the law firms to access legal talent, and that talent is radically homogenous with respect to experience and backgrounds.

If, say, companies were instead working off open sourced documents, then those documents should theoretically benefit from OSS network effects. A stay at home mother in Kansas with an interest in, say, CFAA litigation, should be able to issue pull requests and comments against these documents, develop an online reputation as a niche wonk, and get work that way. And because she’s in Kansas and doesn’t have overhead, she should be able to offer her services above market for her region, but well below market for, say, SF or NYC, increasing competition and driving down the artificially high bubble in those sectors.

But I digress. Simply meant to give a thumbs up on the post. Excited to see an increasing level of conversation around these topics.

I’m coming back to this post because I’m actually trying to figure out how to open-source a set of documents (answer, discovery requests, and affidavit of service for a Minnesota debt buyer lawsuit — here’s a PDF).

Github doesn’t seem like it will work because you can’t really do versioning on Word documents in a user-friendly way. But I can’t use Markdown because the formatting options are so limiting. I cannot even create a proper caption.

HTML could work, but I think using HTML (or Markdown, for that matter) creates a huge barrier to collaboration. For better or for worse, lawyers use Word.

Docracy is no good, either. The editor does not include tables, which means I can’t even format a caption. (I don’t know how it can claim to be a legal forms website when you can’t even format a caption, but whatever.)

Google Docs seems like it might actually be the best option. I’m not thrilled about that, but I guess it’s no more proprietary than Word documents. And it does do collaboration and versioning reasonably well. It’s missing some things I would like to have, like fields with prompts, but I can deal.

Any other ideas? I really need a Github for legal documents, but I’m trying to make do with what’s already out there.

I agree that for better or worse Lawyers use Word – actually we use WordPerfect which I understand some Federal Courts still use but I think it is a step backwards from Word. Even Word is not always compatible with Word since there seems to be about 100 different Word formats.

But it seems like using these proprietary formats for sharing and constructing legal documents is sort of like driving a buggy down a super highway.

Lawyers just need to get beyond these proprietary format and adopt an open source format that is freely shareable and adopted to legal needs. All the attempts at Document Assembly are basically stymied by trying to find a common denominator like rtf or else they have to parse the multiple Word formats.

Markdown or some of the alternative like Asciidoc or ReStructured text might be a start but I agree that they are probably too much for lawyers at present

. However the underlying idea that markdown is just text that can be written in almost any text editor and that the content can be separated from the format, that the content can be freely and easily shared, that it can be put under version control, that with the separation of content from form it can be output on browsers, mobile devices or even paper are all huge advantages

Maybe if lawyers can learn the West Key Number System and Blue Book Citations they can learn Markdown

The problem with the idea that form and content are separate things that can be handled separately (i.e., Markdown “source code” used to generate PDF documents) is that when it comes to some kinds of legal documents (pleadings, for example), form is content. Many court rules and traditions define what a caption must look like, the kinds of fonts that are acceptable, margin widths, and more.

I’m not a fan of Word, but if you don’t include formatting information in the source, you are only solving half the problem. WYSIWYG editors aren’t the only way to do that, but if you want lawyers to contribute (and if you don’t, you will just be creating problems, not solving them), you probably should be thinking in terms of WYSIWYG editors, not text editors and markup languages.

You can say lawyers should get away from Word as easily as you can say regular users should learn to love the command line, but neither is going to happen any time soon.