Hermes Skill (dandacompany) - Setup Guide¶
Source: dandacompany/hermes-skill (2⭐)
Skill: dandacompany/hermes-skill@hermes
Installs: 69
Category: Agent Infrastructure
First Seen: May 3, 2026
A comprehensive Hermes Agent self-hosting and day-to-day operation skill. Covers the entire lifecycle: install, setup, model/provider selection, gateway platforms, Slack channel setup, command discovery, updates, plugins, tools, skills, memory, profiles, MCP, cron, Kanban, delegate_task subagents, logs, dashboards, tutorials, and long-running tmux sessions.
This is the most complete third-party Hermes operations skill on skills.sh. Use it as your primary reference for Hermes Agent administration.
Installation¶
npx skills add dandacompany/hermes-skill@hermes
Or install the full repo:
npx skills add https://github.com/dandacompany/hermes-skill --skill hermes
Prerequisites¶
| Requirement | Details |
|---|---|
| Hermes Agent | Any version. Skill auto-detects Hermes installation. |
| Python 3 | For hermes_check.py script |
| Node.js | For npx skills CLI |
What It Provides¶
1. Dynamic Command Discovery (hermes_check.py)¶
The skill ships with scripts/hermes_check.py - a Python script that introspects the live Hermes installation and returns JSON-structured status. Run at the start of every Hermes task:
python3 <skill-dir>/scripts/hermes_check.py --json
This returns: - Hermes version and installation path - Active profile and configuration - Gateway status - Available models and providers - Installed plugins and tools - MCP server connections - Active cron jobs - Memory/storage statistics
2. Command Map Generation¶
Refresh the local command reference snapshot:
python3 <skill-dir>/scripts/hermes_check.py --write-help <skill-dir>/references/command-map.generated.md
This keeps command documentation current even as Hermes ships updates.
3. Operational Coverage¶
| Area | What the Skill Covers |
|---|---|
| Installation | Fresh install on Linux, macOS, WSL |
| Configuration | config.yaml, profiles, environment variables |
| Models | Provider selection, model routing, fallback chains |
| Gateway | Systemd setup, health checks, death loop recovery |
| Platform Integrations | Slack, Telegram, Discord, 27+ messaging platforms |
| Skills Management | Install, create, update, delete skills |
| Memory | Session DB, Honcho, GBrain, Sibyl-Memory |
| MCP | Server configuration, tool discovery, OAuth |
| Cron | Schedule, manage, and monitor recurring jobs |
| Subagents | delegate_task patterns, parallel dispatch |
| Monitoring | Logs, dashboards, health checks |
| Persistence | Tmux sessions for long-running agent processes |
Quick Start¶
# 1. Install the skill
npx skills add dandacompany/hermes-skill@hermes
# 2. Verify Hermes is healthy
python3 <skill-dir>/scripts/hermes_check.py --json
# 3. Load the skill in your agent
skill_view(name="hermes")
# 4. Follow the skill's guidance for your specific task
Verification¶
After installation, verify the skill loaded correctly:
hermes skills list --marketplace | grep hermes
Then run the health check:
python3 $(find ~/.hermes -name "hermes_check.py" -path "*/dandacompany*" 2>/dev/null | head -1) --json
Expected output: JSON object with version, profile, gateway_status, models, mcp_connections, cron_jobs_count.
Common Pitfalls¶
- Stale command references: Hermes CLI updates frequently. Always refresh with
--write-helpbefore relying on specific command flags. - Python path: The script requires Python 3. On some systems, use
python3explicitly. - Profile mismatch: The skill detects the active profile automatically, but verify with
hermes profile currentif results seem wrong.
Security¶
Related: hermes-agent-framework-setup.md, hermes-agent-helm-setup.md
Integration with CorpusIQ¶
CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.
This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.