FAQ
Common questions about RavenKit and Corvus.
What is Corvus?
Corvus is the in-IDE assistant inside RavenKit. It’s a chat agent with tool-calling: it can read and edit files, run commands, query FIS and vision, create PRDs from text, and sync tickets. It uses your workspace and code to ground answers. You can run it in VS Code (Cmd+Shift+R) or from the terminal with the corvus CLI.
Do I need Ollama?
Ollama is the default and recommended way to run Corvus locally (no data leaves your machine). You can also use Anthropic (Claude) or OpenAI by setting the provider and API key in Settings. Many teams use Ollama for day-to-day and a cloud provider as fallback.
Can I use Claude or OpenAI?
Yes. Set ravenkit.agent.provider to anthropic or openai and add the corresponding API key in Settings. See Configuration.
Where is my data stored?
With Ollama, everything runs locally; your code and chat never leave your machine. With Anthropic or OpenAI, prompts and responses are sent to their APIs per their privacy policies. FIS artifacts (PRDs, ERDs, features) live in your repo under your chosen path (e.g. .features).
How do I install RavenKit?
Install from the VS Code Marketplace (when available), or download the VSIX from GitHub Releases and use Install from VSIX, or build from source. See Get started for step-by-step instructions.
What’s the corvus CLI?
The corvus CLI runs the full RavenKit agent from the terminal—same tools and FIS as in VS Code. Use it in any folder (that folder is the scope) or with --dir. Build the extension first (pnpm compile in apps/ravenkit); then run node dist/agent-cli.js or use an alias. See Troubleshooting for CLI notes.