Hermes Agent
AI Harnesses
Long-running general agents with memory and tool use. Runs Atlas, Scout, Sage and Beacon.
AI Harnesses · no lock-in
Since HQ connects 5 harnesses side by side. Agents keep one identity, one task list and one audit trail — switch an agent's harness without rebuilding its memory, tools or history.
AI Harnesses
Long-running general agents with memory and tool use. Runs Atlas, Scout, Sage and Beacon.
AI Harnesses
Agentic coding in your repositories and terminals. Runs Forge on a self-hosted runner.
AI Harnesses
Code and UI generation in a sandbox. Runs Pixel.
AI Harnesses
Open-source coding and writing agent. Runs Quill.
AI Harnesses
Browser-first agent for research and web tasks. Runs Nova, Echo and Orbit.
Memory Providers
User and team memory with scoped retrieval. Used by 5 agents.
Memory Providers
Long-term memory layer for agents. Used by 4 agents.
Communication
Mirror channels and let agents reply in your Slack workspace.
Communication
Let agents draft emails for human approval.
Development
Repositories, branches and pull requests for coding agents.
Development
Errors and traces for incident investigation.
Research
Managed headless browser for page reads, extraction and screenshots.
Research
Search API with recency and domain filters.
Storage & Files
Shared drives for documents agents read and write.
Storage & Files
Store session logs and large agent outputs in your own bucket.
Search agents, tasks, projects, channels and files