An app store for
your knowledge graph
Install apps into your own Cortex with one click. Connect the places where your knowledge already lives — cloud drives, archives, YouTube — and it flows into your graph automatically, on your schedule, on your infrastructure.
Like an app store —
except everything is yours
Cortex apps don't run in someone else's cloud. They install from the public registry into your own instance and work there — sandboxed, permission-scoped, and checksum-verified from publication to install.
One-click install
Browse the registry inside Cortex (Settings → Apps) or on the web, hit install, done. Updates show up in the same place and never lose your settings.
Your instance, your data
Apps run inside your own Cortex — documents and credentials never leave it. Each app declares exactly what it may access, visible before you install, enforced after.
Set it, forget it
Sync apps keep working with the browser closed: pick folders, pick a schedule, and new or edited files keep flowing into your knowledge graph on their own.
Available today.
Free and open source.
Every app links to its public source repository, and each listing shows exactly what the app may access before you install it.
Nextcloud Sync
Your Nextcloud folders, always in the graph
Pick folders in your Nextcloud and Cortex keeps them in sync — new and edited files flow into your knowledge graph on a schedule, fully automatic, even with the browser closed. Edited files update their existing entry instead of piling up duplicates.
Dropbox Sync
Selected Dropbox folders, synced on a schedule
Connect your Dropbox account in a minute — no app secret, no server setup — choose folders, and Cortex tracks exactly what changed through Dropbox's own change feed. New and updated documents land in your knowledge graph automatically.
Google Drive Sync
Share a folder, and it's in the graph
No OAuth consent screens, no app verification: your instance gets its own service account, and anyone — Workspace or personal Gmail — simply shares Drive folders with its email address, like sharing with a colleague. Regular files sync as-is; Docs, Sheets, and Slides arrive as PDFs.
OneDrive Sync
Your OneDrive folders, tracked by change feed
Connect with a simple device code — no client secret, no redirect servers — and Microsoft Graph's own change feed tells Cortex exactly what changed. New and edited files land in your knowledge graph on a schedule; updated files replace their previous version.
SharePoint Sync
Team libraries, without the sign-in circus
Sync SharePoint document libraries with app-only access to exactly the sites your admin grants — no user sign-in, no consent screens, no tenant-wide permissions. Microsoft Graph's change feeds keep selected libraries and folders flowing into the knowledge graph on a schedule.
WebDAV Sync
One app for a whole world of storage
Sync folders from anything that speaks WebDAV: Nextcloud, ownCloud, Synology and QNAP NAS boxes, Koofr, pCloud, GMX & WEB.DE, Telekom MagentaCLOUD, HiDrive, Infomaniak kDrive, Fastmail, Yandex Disk, Hetzner, Seafile — plus self-hosted servers like rclone and dufs. One protocol, dozens of providers.
Paperless Sync
Your document archive becomes living knowledge
Connect your paperless-ngx archive and every scanned invoice, contract, and letter flows into the knowledge graph — on demand or on a schedule. Cortex ingests the full original documents, so the graph sees layout, tables, and images, not just the OCR text.
YT Transcriber
Videos and whole channels, as searchable text
Paste a video or channel URL and Cortex transcribes, cleans up, and files every video into your knowledge graph. Hours of talks, tutorials, and interviews become queryable knowledge you can ask questions about.
Already part of
every Cortex
Some ingestion paths ship with the platform itself — nothing to install, just connect and go.
Web Import
Built in: harvest the web into your graph
Paste URLs or discover links from any page, and Cortex harvests them into clean markdown — then chunks, embeds, and runs entity extraction just like an upload. Built in and powered by crawl4ai.
Git Integration
Built in: sync your repos into the graph
Connect GitHub, GitLab, or Gitea repositories as a living knowledge source. Cortex ingests your code, docs, and wikis and keeps the graph in sync as files change — and your research agent can act back on the repo by opening pull requests. Built in, with incremental sync.
Apps beyond the instance.
Seamlessly connected.
Not every app lives inside Cortex. Standalone apps are full applications with their own deployment and their own users — connected to your knowledge graph through the REST API with scoped keys. Cortex Chat is the first; anything you can build against 60+ API endpoints can be next.
Cortex Chat
A multi-user chat app for your whole team
The multi-tenant chat app for Cortex. As an admin, provision an account for each team member and assign them to a group scoped to the collections they should see. Every conversation is saved to their own history and follows them across devices, with answers grounded in your knowledge graph — deep research mode, streaming, and inline citations included. No agent to build.
Describe an app.
Your AI assistant builds it.
The whole app system is open. Point any AI coding assistant at the builder skills and describe what you want — the skills teach it to scaffold from the open template, declare the right permissions, and package a working app for your instance. Publishing to the public registry is one pull request away. Every app on this page was built exactly this way.
Prefer to have it done for you? Describe the app you need — a proprietary data source, a custom dashboard, a bespoke workflow — and we build it for your team. Enterprise plans can include custom app development.
Hermes Agent & Cortex.
The recommended harness, with real memory.
Hermes Agent is Nous Research's open-source, self-improving agent harness — our recommended way to run agents on top of Cortex. Install the Cortex skill and Hermes gains your organization's knowledge graph as long-term memory, so it persists, learns, and grows across sessions instead of starting cold.
The Harness
Hermes Agent
Nous Research's open-source, self-improving agent. A closed learning loop — persistent memory, autonomous skill creation, and a deepening model of who you are across sessions. Built to last decades, not demo cycles.
The Memory
Cortex Knowledge Graph
Your organization's documents become a living, traversable knowledge graph. Hermes queries it for grounded context and writes back what it learns — the memory layer that lets the agent genuinely understand your world instead of running stateless.
The Integration
Cortex Skill
Drop the open-standard Cortex skill into Hermes and the connection is instant — no custom glue. The agent reads from and writes to your graph through a single portable SKILL.md, available from the Skills Hub.
An open-source harness and an open memory layer — together, agents that remember, learn, and compound what they know across sessions, on a stack you control.
App questions
Everything about the Cortex app ecosystem.
Small applications that install into your own Cortex instance with one click — the same way you'd install an app on your phone. Most of them connect a place where your knowledge already lives (Dropbox, Nextcloud, YouTube, a document archive) and keep it flowing into your knowledge graph automatically. Web Import and Git Integration are built into the platform itself; the installable apps come from the public Cortex Registry.
Ready to extend your knowledge graph?
Register now and connect your first source — the registry apps are free, open source, and a click away.