This page is for decisions, not raw link lists. Use it to decide where Hermes should live, which provider to start with, and when to reach for MCP, skills, or delegation.
Start here if your goal is the fastest first successful message.
Solo operators, personal assistants, quick mobile access, lightweight ops alerts.
You need heavy server-channel permissions or a team-facing moderation workflow on day one.
Pick this when Hermes needs to live inside an existing community or team server.
Developer groups, internal team bots, shared support channels, role-based coordination.
You only need a private one-to-one assistant and want the shortest setup path.
Best default when you want the most guided Hermes setup path.
First-time Hermes users, users who want fewer provider decisions, users who value the smoothest first run.
You need to compare many hosted models before choosing a long-term provider.
Use this when model choice matters more than setup simplicity.
Users benchmarking providers, switching between vendors, or optimizing for task-specific models.
You want the shortest initial decision path and do not need broad routing yet.
Use local or self-hosted endpoints when privacy and operator control beat convenience.
Ollama, vLLM, air-gapped experiments, sensitive internal workloads.
You need the easiest onboarding and do not want to maintain inference infrastructure.
Use MCP when Hermes needs to talk to external systems with explicit tool boundaries.
Best for filesystems, GitHub, databases, and internal APIs where you want a narrow, reviewable integration surface.
Source: Use MCP with HermesUse skills when the same procedure keeps repeating and should become reusable.
Best for setup checklists, operator runbooks, and repeatable workflows that should not be manually re-explained each session.
Source: RepositoryUse delegation when one context window is no longer enough for the whole job.
Best for code review, parallel research, or isolated subtasks where a fresh context improves output quality.
Source: Delegation guide