The marketplace for intelligent AI agents
Discover, activate, and run AI-powered genes on your Feluda instance.
AI Models
Browse supported AI providers from the Feluda catalog.
Why Feluda
A practical, privacy-conscious way to use AI for real work—without writing code.
No-code workflows
Build repeatable processes in Studio, test them in Workbench, and run them again whenever the same work returns.
Local or cloud models
Connect a supported cloud provider or run a compatible local model on your own computer, depending on your needs.
One trusted package. A complete new capability.
A Feluda Gene bundles the workflows, prompts, tools and resources an AI agent needs to perform a specific job. Enable it once, and the complete capability becomes available to your agent.
Feluda Gene Builder Tool
Approved functions and integrations the agent can use to take action within the permissions you provide.
Built for governance
Use clear steps, review points, and consistent outputs so AI-assisted work stays controlled and auditable.
Schedule and record results
Run tested workflows at the right time and keep useful outputs organised in your Feluda Journal.
Frequently Asked Questions
- No. Feluda is designed for non-developers. You can connect a model, run conversations in Workbench, and build visual workflows in Studio without writing code.
- A Gene is an installable package that adds capabilities to Feluda, such as tools, prompts, workflows, and supporting resources for a specific task.
- Yes. Feluda can work with compatible local models (through supported runtimes) as well as cloud providers. Use local models when you prefer on-device processing.
- MCP servers provide approved tool connections so Feluda can retrieve information or perform permitted actions as part of a workflow, instead of relying on copy/paste.
- Browse the Gene Library, add a Gene to your account, then sync in the Feluda desktop app. After that, its tools or workflows can appear where they’re designed to be used.
- Yes. You can connect compatible local model runtimes and use them inside Workbench and Studio workflows. This allows you to keep more processing on your own machine when that fits your privacy or reliability requirements.