Hermes Agent v0.6.0 lands profiles, MCP server mode, and official Docker container
TL;DR
Hermes Agent v0.6.0 (The Multi-Instance Release) shipped March 30, 2026. Profiles let you run multiple isolated Hermes instances from one install. MCP Server Mode exposes the agent's conversations and sessions to Claude Desktop, Cursor, VS Code, and any MCP client. Official Docker container, two new messaging platforms (Feishu/Lark, WeCom), and an ordered fallback provider chain round out the release.
What changed
What shipped
Hermes Agent v0.6.0 (The Multi-Instance Release) landed March 30, 2026. The headline is that one Hermes install can now run many isolated instances.
Profiles
Each profile has its own config, memory, sessions, skills, and gateway service. Commands:
- •hermes profile create
- •hermes -p [name] to switch
- •hermes profile export / import for sharing
A token-lock mechanism prevents two profiles from sharing a bot credential (so you can't accidentally have two profiles both answering on the same Telegram account).
MCP Server Mode
hermes mcp serve exposes Hermes conversations and sessions to any MCP client. Claude Desktop, Cursor, VS Code, Windsurf: they can all query your Hermes sessions through the standard MCP interface. Supports stdio transport and Streamable HTTP.
This flips Hermes from being a consumer of MCP servers to also being a producer. Your Hermes skill history becomes queryable context for other AI tools you use.
Official Docker container
The Dockerfile supports both CLI mode and gateway mode. Standard docker run pattern; good starting point for self-hosting on a VPS.
Ordered fallback provider chain
Configure a priority list of providers. When the primary errors or is unreachable, Hermes auto-fails over to the next. This is the same pattern OpenRouter offers for model calls, but now at the provider level in Hermes itself.
Two new messaging platforms
- •Feishu / Lark with event subscriptions and interactive cards.
- •WeCom with text, image, voice, and callback verification.
Other wins
- •Slack multi-workspace OAuth so one Hermes install can serve multiple Slack workspaces.
- •Telegram webhook mode for lower-latency message delivery vs polling.
- •Exa search backend alongside Firecrawl and DuckDuckGo as search provider options.
Availability
Upgrade via standard channel. Profiles migration is opt-in (existing single-install configs continue to work as a default profile).
Who this matters for
- Vibe Builder: Profiles let one Hermes install run your personal agent, your work agent, and per-project agents in complete isolation. No more separate installs.
- Basic User: Official Docker container makes self-hosting a 30-second ship. Feishu and WeCom support if you work in the Chinese ecosystem.
- Developer: MCP Server Mode exposes Hermes sessions as queryable context to Claude Desktop, Cursor, VS Code. Ordered fallback provider chain prevents single-provider outages from taking down your agent.
What to watch next
Profiles is the feature that changes what Hermes is good for. Previously, running multiple Hermes agents (one for personal, one for work, one for a specific project) meant multiple installs, separate configs, and version drift. With profiles, it is one install, many isolated instances, all updated together. For anyone running Hermes for more than one purpose, this is the deployment model that actually scales.
MCP Server Mode is the strategic move. By exposing Hermes as an MCP server, your agent's session history and skills become queryable from Claude Desktop, Cursor, VS Code, and any other MCP client. This flips Hermes from being an island (valuable only when you use it directly) to being infrastructure (valuable across every tool you already use). The ecosystem effect compounds over time.
Official Docker removes the single biggest self-hosting friction. Before v0.6.0, deploying Hermes to a VPS involved Python environment setup, systemd service configuration, and dependency pinning. Docker run with the right env vars is a 30-second ship. For anyone running Hermes on a $5 VPS (the canonical deployment), this is the onboarding improvement that matters.
Feishu and WeCom coverage is the Chinese ecosystem play. For teams operating in China or serving Chinese customers, these are table-stakes integrations most Western AI tools have not added. Including them in v0.6.0 rather than deferring to a specialised fork signals Nous Research takes global coverage seriously.
by Harsh Desai
About hermes-agent
View the full hermes-agent page →All hermes-agent updatesMore from hermes-agent
- FeatureHermes Agent v0.4.0 adds OpenAI-compatible API, 6 messaging platforms, and @file context
Hermes Agent v0.4.0 (The Platform Expansion Release) shipped March 24, 2026. It exposes Hermes as a drop-in OpenAI-compatible /v1/chat/completions endpoint, adds 6 new messaging adapters (Signal, DingTalk, SMS via Twilio, Mattermost, Matrix, Webhook) bringing the count to 9, introduces @file and @url context injection, adds 4 new inference providers, and enables gateway prompt caching by default.
- FeatureHermes Agent v0.5.0 expands to 400+ models and adds supply-chain hardening
Hermes Agent v0.5.0 (The Hardening Release) shipped March 28, 2026. Nous Portal now routes to 400+ models through one endpoint. Hugging Face becomes a first-class inference provider. Telegram gets private chat topics with per-topic skill binding. Plugin lifecycle hooks activate. Supply-chain hardening removes a compromised dep, pins version ranges, and adds CI scanning.
- FeatureHermes Agent v0.7.0 adds pluggable memory, credential pools, and Camofox stealth browser
Hermes Agent v0.7.0 (The Resilience Release) dropped April 3, 2026. It opens memory as an extensible plugin system (Honcho as the reference implementation), adds same-provider credential pools with automatic 401 rotation, integrates the Camofox anti-detection browser, and exposes ACP so VS Code, Zed, and JetBrains can register their MCP servers through the agent.