RECAP Project Ideas

RECAP is a free extension for Firefox that strives to provide open access to federal court documents.  The extension seeks to improve judicial transparency by crowdsourcing the collection of court records from PACER—the electronic public access system for the U.S. federal district and bankruptcy courts.  When a RECAP user purchases a document from PACER, the RECAP extension helps her automatically send a copy of that document to the RECAP archive. And RECAP saves its users money by notifying them when documents they’re searching for are already available for free from the public. Thousands of people are actively using the RECAP extension to access over 2 million legal documents. It has been widely reported on in the press (see some quotes here).

Project home:
Project release:

There are many requested features from end users, which are publicly available at RECAP Uservoice page.

Some highly requested features include:
  • Extending the RECAP collection to documents from the U.S. appeals courts.

    RECAP currently only supports the district and bankruptcy PACER sites.  We want to support the appellate court PACER sites, but the interface is quite different than the district/bankruptcy versions.  This project would modify both the extension and the extension server to handle document uploads from the appellate sites.
  • Google Chrome and Internet Explorer versions of the RECAP extension.

    Many potential users of RECAP do not have the ability to install and use Firefox in their work environments or prefer to use another browser.  Having both Chrome and IE versions of the extension will significantly expand our user base, enabling many more users to contribute to the RECAP document repository. (see

  • Capturing "free look" documents that are provided to attorneys in a case.

    After a document is filed electronically to the court, the attorneys in the case receive an e-mail notification with a link to a "free look" for the document.  By adding functionality to RECAP to automatically capture "free look" documents, the document repository would stay up-to-date with new documents for ongoing cases. (see these two items on our uservoice page)
  •  An interface for attorneys to submit court documents they have stored locally.
Many law firms and attorneys have large local archives of PACER documents stored on their own systems that they are willing to donate to the RECAP document repository.  We need to develop an interface to easily accept these documents, categorize them, and upload them to our repository. (uservoice item)
  • A better browsing interface for the RECAP document repository.
We've built a nice website that allows users without a PACER account to search and browse the document repository.  The site is currently in private beta, and there are plenty of opportunities to add useful features, such as collaborative annotation or case summarization, as we take the site public.
  • More advanced scraping of metadata contained in court dockets.
We get hundreds of dockets from RECAP users every day, and we parse these dockets for useful metadata.  Better scraping of these dockets will allow us to capture much more information about court documents and make RECAP more useful both to ordinary users and researchers.
  • Automatically crawl free opinion reports.
Most PACER instances publish a report of all new opinions along with free download of those documents.  Unfortunately, these reports are in an obscure location on PACER and the interface is not very usable.  We could script a crawler to pull this data out and stick it into the RECAP database for more productive use.

Skills you might need, depending on the project idea:
  • Javascript / Mozilla extensions experience (for the client extension)
  • Python / Django experience (for the extension server)
  • MySQL
  • Haystack / Solr / Lucene (for the search front-end)
  • Regexp parsing / BeautifulSoup experience (for docket scraping/parsing improvements)
  • XML/XSLT manipulation (for display transformations)
  • AJAX/JSON experience for client-server interactions.
The codebase: