OpenAI builds and deploys advanced AI models like GPT-4o for autonomous agents and workflows.
Krelvan
0Popularity
About Krelvan
Krelvan is a self-hosted, open-source platform that converts plain English descriptions into working agent systems capable of performing tasks across integrated tools. Users define an outcome and the system generates a validated agent graph that executes locally without vendor lock-in. The platform records every step of execution, allowing users to replay, inspect, and audit runs in detail. Agents can be scheduled to run unattended or paused for human approval at specific steps, with outputs delivered through built-in channels such as email, Slack, or Telegram. The system supports multiple LLM providers including local models via Ollama and integrates with a Git-based registry for sharing and reviewing agent templates. All infrastructure runs on the user’s own machine, with data and execution logs remaining under local control.
Key features
- Plain English to agent conversion
- Step-by-step execution replay
- Multi-LLM provider support
- Human-in-the-loop approval gates
- Built-in Agent Inbox for output delivery
- Git-based registry for agent templates
- Scheduled unattended runs
- OS-process sandboxing for untrusted plugins
Use cases
- Automating repetitive research tasks
- Monitoring and alerting via scheduled agents
- Drafting and reviewing documents with approval steps
Pros
- Converts plain English goals into executable agent systems
- Full execution history with step-by-step replay capability
- Supports multiple LLM providers including local models
- Built-in delivery to email, Slack, Telegram, and webhooks
- Self-hosted with no vendor lock-in
Cons
- Requires local installation and self-hosting
- Early release with some components marked as untested
- No hosted SaaS option available
Frequently asked questions about Krelvan
What is Krelvan?
Krelvan is a self-hosted, open-source platform that converts plain English descriptions into working agent systems capable of performing tasks across integrated tools. It generates a validated agent graph that executes locally without vendor lock-in.
Who is Krelvan for?
Krelvan suits users who want full control over their agent infrastructure, including developers, researchers, and teams prioritizing data privacy and auditability. It is designed for those comfortable running local or self-hosted systems.
How does Krelvan handle pricing?
Krelvan is open-source and self-hosted, meaning users download and run the software on their own infrastructure without subscription fees. Optional marketplace templates or connectors may have their own licensing terms.
What integrations does Krelvan support?
Krelvan integrates with multiple LLM providers, including local models via Ollama, and supports delivery channels like email, Slack, Telegram, WhatsApp, and webhooks. It also offers a Git-based registry for sharing and reviewing agent templates.
Can Krelvan run agents without human intervention?
Yes, Krelvan allows agents to be scheduled to run unattended via cron or interval schedules. However, users can configure specific steps to pause for human approval or review before execution.
How do I get started with Krelvan?
Users can start by installing Krelvan locally using a version-pinned command, Docker, or an inspectable tarball. The platform boots an API and web UI on localhost:3100, enabling users to describe goals and generate agents immediately.
