Observal provides an open‑source registry and analytics platform for AI coding agents, letting developers upload, version, and track components such as skills, MCP servers, hooks, prompts, and sandboxes. These components can be packaged into portable agents and installed across multiple IDEs and tools—including Claude Code, Cursor, Kiro, Gemini CLI, VS Code, and GitHub Copilot—with a single command. The platform also offers analytics, insights, and session trace replay to show which agents and tools are delivering value.
Funding
Funding not disclosed
Founders
Product
Problem
Developers of AI coding agents lack a centralized way to manage, version, and distribute reusable components such as skills, prompts, hooks, and sandbox environments, leading to fragmented workflows and difficulty tracking usage across different IDEs.
Solution
Observal offers an open‑source registry that lets developers upload and version AI agent components in a single catalog. Components can be assembled into portable agents that are installed across multiple IDEs—including Claude Code, Cursor, Kiro, Gemini CLI, VS Code, and GitHub Copilot—with a single command. Built‑in analytics provide insight into which agents and components are most effective, while session traces record every tool call and token for debugging and optimization. The platform is self‑hosted under an AGPL‑3.0 license, allowing organizations to run the service on their own infrastructure.
Target Audience
Primary users are developers and engineering teams building AI coding agents who need a unified component marketplace and observability across multiple development environments.
Features
- Component Registry for uploading, versioning, and tracking skills, MCP servers, hooks, prompts, and sandboxes
- Agent Registry that packages selected components into portable agents installable with one command
- Cross‑IDE support for Claude Code, Cursor, Kiro, Gemini CLI, VS Code, and GitHub Copilot
- Analytics dashboard showing usage metrics and performance of agents, skills, and tools
- Session trace replay of each tool call and token to aid debugging and optimization
- Open‑source, self‑hosted deployment under AGPL‑3.0 with full access to source code