An agent in Hermes is not just a model responding to a prompt. It is a complete working identity — a model, a system prompt, tools it can call, memory it carries forward, and a configuration that ties it all together. When you define an agent, you are defining a role: who it is, what it knows, what it can do, and how it behaves.
Think of it like hiring for a position. You do not just hire "someone smart." You hire a specialist with a defined role, a set of responsibilities, access to specific tools, and instructions about how to do the work. The model is the brain; everything else — profile, skills, memory, toolsets, gateway, cron — is the job description, the desk, the filing cabinet, and the schedule.
A profile is an isolated identity. It holds its own system prompt, configuration, memory, skills, and session history. Two profiles on the same machine are two separate agents — they do not share context, they do not see each other's files, and they do not interfere with each other. This matters because you want agents with defined roles, not generalists that try to do everything. Separate profiles with narrow instructions produce sharper results.
In your SEO workflow, you need two distinct roles: an SEO manager holding strategy context, brand voice, and publishing standards; and a research specialist focused on keyword research and structured data. These two profiles are where the agent team starts — one directing the work, one doing the research.
Profile isolation means the SEO manager's strategy decisions live in the manager's memory — not in the research specialist's. Each agent sees only what its role requires.
A skill is a reusable procedure stored as a document (SKILL.md format) and loaded automatically when relevant to the task at hand. Think of skills as standard operating procedures — a restaurant has one for opening, one for closing, and one for a health inspection. Staff follow the one that applies; agent skills work the same way.
In your SEO workflow, the research specialist has a skill for keyword research: which tools to call, how to filter results, what format to return. The SEO manager has a skill for content briefs: how to turn keyword data into a structured brief, what sections to include, what style rules to follow. Over time, agents can also save useful patterns into skills — not independently, but by following your instructions to persist what works.
Skills are loaded on demand, not all at once. The agent picks the ones that match the current task — keeping context focused and costs controlled.
Memory is what separates a persistent agent from a disposable chat. The built-in system uses two primary documents per profile: a general memory file for facts the agent has learned, and a user preferences file for your identity and style rules. Both load into context at the start of every session. Optional external providers (like Honcho or Mem0) add semantic search and richer recall — but the built-in system is enough to start.
In your SEO workflow, memory is where compounding value lives. Tell the SEO manager to always write in a specific tone, and that rule persists across sessions. If the research specialist discovers a keyword source that works better for your niche, it saves that finding. The longer you work with the agent, the less setup you need — because it carries forward what you taught it.
Memory is not perfect recall. You control what persists by giving clear instructions about what to save and what to discard.
A toolset is a group of related tools with shared dependencies. Hermes includes over 70 built-in tools organized into toolsets — file operations, web search, shell execution, and more. You enable or disable toolsets per profile, so each agent only has access to the tools it needs. This is how an agent moves from generating text to actually doing things: a chat model can suggest you search the web; a Hermes agent with the web search toolset enabled performs the search itself.
In your SEO workflow, the research specialist needs web search and file writing toolsets — to find keyword data and save results. The SEO manager needs file reading and writing toolsets — to read research results and draft content. Narrow tool access keeps each agent focused on its role.
The gateway connects your Hermes agent to over 20 messaging platforms — Telegram, Discord, Slack, WhatsApp, Signal, Email, and more. Instead of opening a terminal to talk to your agent, you send a message on the platform you already use. The gateway translates your message into an agent prompt and the agent's response back into a readable message.
This removes a real barrier: if you have to open a terminal and start a session every time, you will use the agent less. The gateway lets you message your agent the same way you message a colleague — from your phone, from Slack, from Telegram. In your SEO workflow, you can ask the research specialist for keyword ideas from Slack and get results in the same thread, all without leaving your messaging app.
Cron is the built-in scheduler. It runs agent tasks on a timed interval — every 30 minutes, every Monday at 9am, or any schedule you define. You can schedule tasks using natural language or standard cron expressions. Results get delivered wherever you want: back to your messaging platform, saved to a file, or chained into the next task. Cron also supports a no-agent mode for lightweight checks and data-fetching jobs that do not need AI reasoning.
In your SEO workflow, cron is how you go from asking the research specialist for keywords manually to having it run a keyword scan every Monday morning and deliver results to your Slack channel. You do not need to remember to ask — the schedule handles the asking, and you handle the reviewing.
Subagents are temporary agents spawned for specific tasks. They start with fresh context, have access to the same tools as the parent profile, and return their results when done — useful for parallel work without contaminating the main agent's context. Kanban is the coordination system for multi-agent work: a shared task board where agents claim tasks, report progress with heartbeats, and hand off completed work. If an agent stops responding, the system detects it and can reclaim the task.
In your SEO workflow, these come into play once the team grows beyond two profiles. For now, the SEO manager directs the research specialist through direct task assignment — no kanban needed. But when you add a SERP analyst, a content writer, and a reviewer, coordination becomes necessary. Later chapters will show how.
Your SEO workflow combines these components into a working team: the SEO manager profile holds strategy context and uses file tools to draft briefs; the research specialist profile uses web search and a keyword research skill to gather data. Memory lets both carry forward what they learn — your tone preferences, which sources work best. The gateway lets you reach them from Slack. Cron will later schedule the Monday keyword scans automatically. And subagents and kanban will coordinate the team once it grows beyond two roles.
The model is the engine inside each profile, but the surrounding components — profile, skills, memory, toolsets, gateway, cron — are what turn that engine into a useful worker. Every addition in later chapters builds on these same components. Understanding them now makes everything that follows easier to reason about.
Think about a recurring task in your own work. If you were to design two agent profiles for it, what would each profile's role be? What skills would each need? What memory would each carry forward?