Hi Lanes team 👋
I'm from the Hermes Workspace project (https://github.com/jack-wz/hermes-platform) — an open-source AI coworker platform with an 8-skill registry, real subprocess execution, and multi-channel connector.
The idea: Hermes could serve as the AI backend execution layer for Lanes — providing agent lifecycle management, skill orchestration, and coworker dispatching while Lanes handles the browser automation and observability frontend.
What Hermes brings:
- Coworker engine: 9 pre-built AI coworkers (morning-brief, signal-scanner, content-writer, etc.)
- Real subprocess execution via Dashboard API (no stubs)
- SKILL.md registry with 8 skills (scoring A-F, sandbox verified)
- Multi-channel delivery (Feishu, Slack, Discord, WhatsApp)
- Kanban board with dependency management
- MCP Gateway for model routing
What a Lanes+Hermes integration could look like:
- Lanes captures a browser workflow → Hermes coworker picks it up
- Lanes detects an anomaly → Hermes creates a kanban task + dispatches investigation
- Shared memory between Lanes' observations and Hermes' reasoning
Open to a quick chat or issue thread to explore if this makes sense. No pressure — just reaching out because the architecture alignment feels natural.
Cheers!
Hi Lanes team 👋
I'm from the Hermes Workspace project (https://github.com/jack-wz/hermes-platform) — an open-source AI coworker platform with an 8-skill registry, real subprocess execution, and multi-channel connector.
The idea: Hermes could serve as the AI backend execution layer for Lanes — providing agent lifecycle management, skill orchestration, and coworker dispatching while Lanes handles the browser automation and observability frontend.
What Hermes brings:
What a Lanes+Hermes integration could look like:
Open to a quick chat or issue thread to explore if this makes sense. No pressure — just reaching out because the architecture alignment feels natural.
Cheers!