DOCS
Getting started
Everything you need to install Nymiria, connect your AI providers, and start orchestrating.
Installation
macOS
$
curl -fsSL https://nymiria.com/install.sh | bashInstalls via .dmg. Requires macOS 12 or later.
Windows
$
irm https://nymiria.com/install.ps1 | iexRuns the NSIS silent installer. Requires PowerShell 5.1+.
Linux
$
curl -fsSL https://nymiria.com/install.sh | bashInstalls as AppImage on all distros — supports in-app auto-update.
Connecting AI providers
- 1Open Nymiria and go to Setup → Providers.
- 2Click Add key next to any provider (Anthropic, OpenAI, Google, etc.).
- 3Paste your API key — it's stored in your OS keychain immediately.
- 4Select the provider in any channel to start using it.
Local inference (offline mode)
- 1Install Ollama and pull a model:
ollama pull llama3 - 2In Nymiria, go to Setup → Providers and enable Local Inference.
- 3Flip the mode toggle in the channel toolbar — a status badge confirms zero network usage.
- 4Supported backends: Ollama, llama.cpp. Any model your hardware can run.
Channels & workspaces
- 1Each channel is an isolated conversation with its own provider, memory, and context.
- 2Workspace shows your files and lets you attach context to any message.
- 3Projects group channels, memory, and Kanban boards around a specific codebase or goal.
- 4Switch between channels without losing context — each resumes exactly where you left off.
Automations & crons
- 1Type
/cronin any channel to schedule a recurring task. - 2Example:
/cron Briefing | 0 9 * * * | Summarize my Kanban board - 3Crons run against whichever provider the channel is configured to use.
- 4Manage active crons from Setup → Skills.
Keeping Nymiria updated
- 1Nymiria checks for updates automatically on launch.
- 2When an update is available, a badge appears in the bottom status bar.
- 3Click it to download and install — no terminal required.
- 4You can also re-run the install one-liner to update manually.