In active development since 2015, MemoRekall is a free and open source web app for the documentation and preservation of time-based media. The app is an interface that allows for the redocumenarisation of a primary video resource: the user provides a Youtube or Vimeo video to which they can append intra- and inter-documentary, multi-modal annotations (links, files, other MemoRekall projects etc.). The arrangement of these documents creates a capsule which can be shared and embedded. MemoRekall is a powerful yet simple tool that is already used by a wide range of users: researchers, artists, cultural institutions and mediators, teachers and students. MemoRekall is currently in a new phase of development with the primary goal of rendering it compatible with IIIF. This new version is developed as plugins and patches for the widely-used Mirador viewer and takes full advantage of the latest developments of IIIF’s Presentation API. Our source code is available at https://gitlab.tetras-libre.fr/iiif/ and builds upon a previous Mirador fork supporting video from Tokyo University (https://github.com/2SC1815J/mirador), as well as an annotation plugin for images that we extended to support video (https://github.com/ProjectMirador/mirador-annotations). Our contributions include the update of those projects and their extension to support the annotation of a manifest with another manifest. We propose: A fully functional video player and annotator within Mirador built upon existing work mentioned above. The creation of manifest networks: each annotation can refer to another manifest, allowing the creation of complex networks of multi-modal content. Leaning on IIIF’s interoperability capabilities, this also means that networks can be constituted from a number of different sources. Dynamic network navigation in Mirador: annotations linked to other manifests can be opened directly in the same instance of Mirador, allowing for a seamless navigation experience. We also aim to ensure full retrocompatibility with existing MemoRekall projects in order to support our current user-base and encourage a dialogue between them and the IIIF community. This work has been driven by a concrete case study in the context of the Horizon 2020 COESO project. A pilot manifest network was constructed around this project and served as the basis for our working prototype. It has allowed us to follow a practice-led design approach, and has brought to light a number of issues that will be important going forth. Notably, we intend to adopt a git-inspired version control model, allowing for different versions of manifest networks and cumulative and comparative work between different researchers. This addresses a fundamental issue in SHS analysis where each researcher can propose radically different ideas on the same set of ressources. Leaning on IIIF, MemoRekall wishes to propose an environment where these perspectives co-exist in a dynamic and intelligible way. This will also pave the way for the integration of AI- and ML-driven corpus analysis tools which shall be the subject of our next phase of development.