maestro-flow
View on GitHubIntent-driven workflow orchestration for multi-agent AI development — adaptive lifecycle engine, self-reinforcing knowledge graph, and visual dashboard for Claude Code, Gemini, Codex & more
Maestro-Flow is a TypeScript CLI and MCP server for orchestrating multi-agent software development. It routes natural-language goals through adaptive workflows, coordinates multiple agent backends, and stores reusable project knowledge in a graph.
Use Cases
Coordinate multiple coding agents across development tasksPlan and execute software changes from natural-language intentRun adaptive review, testing, and repair workflowsDelegate work across Claude Code, Codex, Gemini, and other agent CLIsPersist project knowledge and inject relevant context into later runsRun long-lived autonomous development and debugging loops
Built With
- Language
- TypeScript
- Frameworks
- MCP SDK · React · Hono · Vite
Tags
multi-agent · workflow-orchestration · AI coding · adaptive workflows · knowledge graph · MCP · CLI · agent delegation · parallel execution · Claude Code · Codex · Gemini