Questions about Krelvan
What is Krelvan?
Krelvan is an open-source, self-hosted software platform that lets users build and run AI agents on their own machines by describing a desired outcome in plain English. It plans the agent, displays it as a visual canvas that functions as the runtime, and executes it locally without requiring a hosted service or an account. The system supports extending agents through an open Git registry of templates, marketplace entries, and MCP connectors, and can use multiple LLM providers or run fully locally.
How does the canvas function as the runtime?
The platform allows users to review the plan before anything runs, ensuring control over execution. Every node in the visual canvas is a real step, and every edge is a real branch, with the entire run replayed over the same recorded run so the graph is the agent itself. This approach ensures transparency and allows users to see exactly what executed at every step.
What output channels does Krelvan support?
Krelvan provides built-in channels for delivering outputs, including the Agent Inbox, email, Telegram, Slack, webhooks, and other configured connectors. Users can deliver results without moving execution to the website, keeping everything within their local infrastructure. These channels allow agents to reach users where they live, integrating seamlessly with existing communication tools.
How does the registry and agent building process work?
The platform includes a registry/index.json that loads inspectable registry agents, allowing users to start with a goal and run it on their own machine. Users can describe what they want done, and Krelvan plans the agent, shows it to them, and runs it locally. This process enables building agents like a price watcher, support bot, personal advisor, or research digest.
How does Krelvan differ from raw agent frameworks?
Krelvan offers a framework that provides an integrated, inspectable implementation you can run and evaluate, rather than just giving you parts to wire yourself. It handles the graph, runner, retries, and human-in-the-loop gates, allowing you to focus more on domain logic. The system also includes failure diagnosis and retry capabilities, where it reads full history to diagnose issues and drafts corrected agents.
What are the licensing and deployment options?
The platform is Apache-2.0 open-source, allowing users to self-host it on their own machine for themselves, their team, or their clients. Users can extend the system and own what they build, with no vendor lock-in. This flexibility ensures that the infrastructure underneath supports full run history, failure diagnosis, human control, and process isolation.
What happens on the first start?
On the first start, the terminal prints a one-time setup token and the local URL. Users open the URL, paste the token, and create the admin account for this installation. The admin account is local to this installation, not a krelvan.com account, and the public website does not receive credentials, agents, or run data.
How do I manage installation maintenance?
Users can stop, restart, back up, or upgrade Krelvan by using Ctrl-C to stop the API and web UI, then running the same versioned command again to restart. Backing up the entire data directory before changing versions is recommended, including identity and secret-key files. Container installs keep the same state in the krelvan-data volume.
Do I need an LLM API key to try Krelvan?
An LLM API key is not required just to try Krelvan, as the platform can run fully locally or use multiple LLM providers. This flexibility allows users to test the system without needing external API keys initially. The system supports various LLM providers, ensuring compatibility with different models.
Who is Krelvan designed for?
Krelvan is designed for users who want to own their agents and keep a complete record of every step. It suits those seeking transparent, customizable agent systems that run on their own infrastructure. The platform is aimed at users who value self-hosting and open-source solutions for building and running AI agents.
How does the run history feature work?
The platform's visual canvas allows users to scrub back through any run, step by step, providing a pure read of the execution record. This feature enables detailed inspection of agent behavior and ensures that what is seen is exactly what executed. Users can review the plan before execution, maintaining full control over the process.
What marketplace features does Krelvan offer?
Krelvan includes a marketplace for agent templates and MCP connectors, allowing users to extend functionality through an open Git registry. Users can publish their own inspectable entry through the registry review process, contributing to the community. This ecosystem supports collaboration and innovation in agent development.
How does Krelvan handle LLM providers?
The platform supports multiple LLM providers, allowing users to choose the model that best fits their needs. Users can also run agents fully locally, providing flexibility in deployment and data privacy. This multi-provider support ensures that Krelvan can adapt to various LLM capabilities and requirements.
What is the official website of Krelvan?
The official website of Krelvan is krelvan.com.
Does Krelvan have reviews on Trustburn?
Krelvan has no reviews on Trustburn yet. Anyone who has used it can write the first review on its Trustburn page.
Post your question publicly — other visitors, and whoever manages this page, can answer. It will be visible on this page.
Want to hear about this company again? Follow it and we’ll email you when a new review appears.
Based on reviews, listing data and the company's website. Verify anything important with the company directly.
Questions & Answers
Does the core platform license include unlimited local agent execution, or are advanced connectors charged separately?
Trustburn Team suggests a starter question on new profiles — the business hasn’t answered yet.
Sign in to post
Your text is saved — you'll come right back to post.
Don't have an account? Create one
News & Press
No news yet. This is where the business shares its own updates and press.
About Krelvan
Krelvan is an open-source, self-hosted software platform that lets users build and run AI agents on their own machines. Users describe a desired outcome in plain English, and the software plans the agent, displays it as a visual canvas that also functions as the runtime, and executes it locally. Agents can be extended through an open Git registry of templates, marketplace entries, and MCP connectors, and they can use multiple LLM providers or run fully locally. Outputs can be delivered through built-in channels such as an Agent Inbox, email, Telegram, Slack, webhooks, and other configured connectors.
- Website
- krelvan.com
- [email protected]

