Roadmap

Firefly evolves around one principle: stabilize core scheduling semantics first, then add platform capabilities through independent modules.

Near Term

• Improve Admin API request and response examples
• Add more Spring Boot Starter examples
• Improve execution history and troubleshooting in Admin UI
• Expand Prometheus alert templates
• Add compatibility tests for external plugins

Mid Term

• tracing plugin
• more Executor transports
• import and export for jobs
• cluster maintenance tools
• schema migration tooling improvements

Long-Term Principles

• Do not put web, storage, monitoring, or transport protocols into scheduler-core
• Do not let default time zones define job semantics
• Do not generate unfenced executions when the database is unavailable
• Do not re-couple Admin UI and Java API