[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-75967":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":15,"stars7d":16,"stars30d":17,"stars90d":14,"forks30d":14,"starsTrendScore":18,"compositeScore":19,"rankGlobal":9,"rankLanguage":9,"license":20,"archived":21,"fork":21,"defaultBranch":22,"hasWiki":23,"hasPages":21,"topics":24,"createdAt":9,"pushedAt":9,"updatedAt":30,"readmeContent":31,"aiSummary":32,"trendingCount":14,"starSnapshotCount":14,"syncStatus":33,"lastSyncTime":34,"discoverSource":35},75967,"hermes-agent-control-room","shannhk\u002Fhermes-agent-control-room","shannhk","Control Room-first template for managing Hermes agents from one VPS agent to specialist teams and orchestrated workflows",null,"Shell",538,89,266,0,5,18,207,20,9.86,"MIT License",false,"main",true,[25,26,27,28,29],"agent-control-room","ai-agents","hermes-agent","multi-agent","vps","2026-06-12 02:03:38","# Hermes Agent Control Room\n\n![Hermes Agent Control Room](assets\u002Fhermes-agent-control-room.png)\n\nA public template for setting up an **Agent Control Room** first, then scaling from one Hermes agent to direct specialists, orchestrated teams, and automated workflows.\n\nThe Agent Control Room is a sidecar repo\u002Ffolder that documents and governs your Hermes agents. It is **not** an agent itself. It is the system map, operating manual, registry, runbook library, and recovery notebook for the agents you run.\n\nIt gives you a clean path from:\n\n```text\none agent -> direct specialists -> orchestrator -> automated agent team\n```\n\n## About\n\nHermes Agent Control Room is a starter kit for people who want to run Hermes agents like an operating system instead of a pile of disconnected bots.\n\nThe repo gives you:\n\n- A control-plane folder structure for documenting agents.\n- Templates for agent runbooks, Docker notes, secret maps, and backups.\n- A level-based architecture for growing from one agent to a specialist team.\n- A task bus pattern for orchestrator-to-specialist delegation.\n- Bundled setup and operations skills an agent can use to build or manage the system.\n\nThe key idea is simple: **set up the Control Room first, then plug agents into it.**\n\n## Core Idea\n\n```text\nCreate a VPS or choose an existing one.\nBootstrap the Agent Control Room.\nRegister one Hermes agent.\nAdd direct specialists when roles become clear.\nAdd an orchestrator when you want one front door.\nAutomate only after the manual system works.\n```\n\nThe Control Room sits on the side as the control plane. You can use it directly, talk directly to any agent, or talk to an orchestrator that delegates to specialists.\n\n```text\nAgent Control Room = side control plane\nOrchestrator       = optional manager\u002Ffront-door agent\nSpecialists        = focused Hermes agents with role-specific tools\nTask Bus           = handoff desk between orchestrator and specialists\nYou                = owner\u002Foperator with direct access to everything\n```\n\n## Full System Shape\n\n```mermaid\nflowchart LR\n  user[\"You \u002F Operator\"]\n\n  control[\"Agent Control Room\u003Cbr\u002F>\u003Ccode>\u002Froot\u002Fagent-control-room\u003C\u002Fcode>\u003Cbr\u002F>\u003Cbr\u002F>side control plane\u003Cbr\u002F>docs \u002F rules \u002F registry\u003Cbr\u002F>ports \u002F env maps\u003Cbr\u002F>runbooks \u002F backups\"]\n\n  orch[\"hermes-orchestrator\u003Cbr\u002F>\u003Cbr\u002F>optional front door\u003Cbr\u002F>delegation \u002F synthesis\"]\n\n  bus[\"Agent Task Bus\u003Cbr\u002F>\u003Ccode>\u002Fsrv\u002Fagent-bus\u003C\u002Fcode>\u003Cbr\u002F>\u003Cbr\u002F>inbox \u002F working\u003Cbr\u002F>outbox \u002F archive\"]\n\n  life[\"hermes-life\u003Cbr\u002F>personal agent\"]\n  seo[\"hermes-seo\u003Cbr\u002F>SEO specialist\"]\n  dev[\"hermes-dev\u003Cbr\u002F>code \u002F site work\"]\n  cmo[\"hermes-cmo\u003Cbr\u002F>marketing\"]\n  ops[\"hermes-ops\u003Cbr\u002F>VPS \u002F backups \u002F security\"]\n\n  user -->|\"control path\u003Cbr\u002F>edit docs \u002F rules\"| control\n  user -->|\"orchestrated path\"| orch\n  user -->|\"direct path\"| life\n  user -->|\"direct path\"| seo\n  user -->|\"direct path\"| dev\n  user -->|\"direct path\"| cmo\n\n  control -. \"defines \u002F documents \u002F governs\" .-> orch\n  control -. \"defines \u002F documents \u002F governs\" .-> life\n  control -. \"defines \u002F documents \u002F governs\" .-> seo\n  control -. \"defines \u002F documents \u002F governs\" .-> dev\n  control -. \"defines \u002F documents \u002F governs\" .-> cmo\n  control -. \"defines \u002F documents \u002F governs\" .-> ops\n\n  orch -->|\"routes tasks\"| bus\n  bus --> seo\n  bus --> dev\n  bus --> cmo\n  bus --> ops\n  seo -->|\"results\"| bus\n  dev -->|\"results\"| bus\n  cmo -->|\"results\"| bus\n  ops -->|\"results\"| bus\n  bus -->|\"summaries \u002F artifacts\"| orch\n  orch -->|\"final synthesis\"| user\n```\n\n## Access Paths\n\nYou are never locked into one workflow.\n\n```text\nControl path:\n  You -> Agent Control Room\n\nDirect path:\n  You -> hermes-seo\n  You -> hermes-dev\n  You -> hermes-cmo\n\nOrchestrated path:\n  You -> hermes-orchestrator -> Agent Task Bus -> Specialists -> You\n```\n\n## Architecture Levels\n\n### Level 1: Agent Control Room + One Agent\n\nSet up the Control Room and register one Hermes agent.\n\nBest for:\n\n- One personal Hermes agent\n- VPS setup documentation\n- Docker migration planning\n- Keeping runbooks and secret maps organized\n\nYou do not need an orchestrator or task bus yet.\n\n### Level 2: Direct Specialist Agents\n\nAdd multiple role-specific Hermes agents and talk to them directly.\n\nExamples:\n\n- `hermes-life`\n- `hermes-seo`\n- `hermes-dev`\n- `hermes-cmo`\n- `hermes-ops`\n\nThe Control Room documents all of them. You choose which agent to talk to.\n\n### Level 3: Orchestrator + Specialists\n\nAdd `hermes-orchestrator` as an optional front door. You can still talk directly to specialists, but the orchestrator can route and synthesize work.\n\nThe orchestrator follows the Control Room. It should not become a giant agent with every credential.\n\n### Level 4: Automated Agent Team\n\nAdd recurring workflows, audits, backup checks, task routing, and optional direct gateway\u002FAPI calls.\n\nOnly add automation after the manual workflow works.\n\n## Bundled Skills\n\nThis repo includes skills that can be linked into Claude Code or adapted for Hermes.\n\n```text\ncreate-vps\n  Create a fresh Hetzner VPS, SSH key, SSH alias, and local provisioning folder.\n\nsetup-control-room\n  Bootstrap an SSH-accessible VPS with Node, Claude Code, Codex, Docker,\n  Hermes Agent, and this Control Room template.\n\nagent-control-room\n  Manage the Control Room docs and agent folders.\n\nagent-task-router\n  Route tasks from an orchestrator to specialists through a task bus.\n\nagent-registry-manager\n  Maintain the agent registry.\n\nagent-backup-manager\n  Design and audit per-agent backups without committing secrets.\n\nagent-security-auditor\n  Check ports, dashboards, SSH, Docker, secret placement, and key scope.\n\nagent-team-cron-planner\n  Plan recurring multi-agent workflows after manual workflows work.\n```\n\n## Suggested Folder Structure\n\n```text\nagent-control-room\u002F\n  README.md\n  agents\u002F\n    .gitkeep\n  docs\u002F\n    architecture.md\n    levels.md\n    naming.md\n    security.md\n    task-bus.md\n    orchestrator.md\n    starter-guide.md\n  shared\u002F\n    api-keys-sop.md\n    commands.md\n    security.md\n  templates\u002F\n    agent\u002F\n      inventory.md\n      docker.md\n      env-map.md\n      runbook.md\n      backup.md\n    docker\u002F\n      docker-compose.agent.yml\n      docker-compose.orchestrator.yml\n    task-bus\u002F\n      agents.yaml\n      task-template.md\n      result-template.md\n  skills\u002F\n    create-vps\u002F\n    setup-control-room\u002F\n    agent-control-room\u002F\n    agent-task-router\u002F\n    agent-registry-manager\u002F\n    agent-backup-manager\u002F\n    agent-security-auditor\u002F\n    agent-team-cron-planner\u002F\n  examples\u002F\n    level-1-control-room-one-agent\u002F\n    level-2-direct-specialists\u002F\n    level-3-orchestrator-specialists\u002F\n    level-4-automated-team\u002F\n```\n\n## Setup\n\nThere are three ways to use this repo.\n\n### Option A: Point An Agent At This Repo\n\nThis repo is designed to be agent-readable.\n\nIf your agent can read a GitHub repo or a local clone, point it here and ask:\n\n```text\nRead this repo and help me set up an Agent Control Room.\nStart with docs\u002Fstarter-guide.md and the setup-control-room skill.\n```\n\nIf the bundled skills are available to the agent, you can be more direct:\n\n```text\nUse setup-control-room to bootstrap my VPS.\n```\n\nOr, if you need a new Hetzner server first:\n\n```text\nUse create-vps, then chain into setup-control-room.\n```\n\nThe intended agent flow is:\n\n```text\ncreate-vps\n  -> creates a Hetzner VPS, SSH key, and SSH alias\n\nsetup-control-room\n  -> installs tooling and clones this repo onto the VPS\n\nagent-control-room\n  -> helps register and manage agents inside the Control Room\n```\n\nImportant: the repo does not magically run code when opened. It gives your agent the setup instructions, templates, and skills. You still ask the agent to run the setup flow.\n\n### Option B: Manual Setup On An Existing VPS\n\nUse this if you already have an Ubuntu\u002FDebian VPS you can SSH into.\n\nSSH in:\n\n```bash\nssh root@YOUR_SERVER\n```\n\nClone the Control Room:\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Fshannhk\u002Fhermes-agent-control-room.git \u002Froot\u002Fagent-control-room\ncd \u002Froot\u002Fagent-control-room\n```\n\nRead the starter guide:\n\n```bash\ncat docs\u002Fstarter-guide.md\n```\n\nRegister your first agent:\n\n```bash\nmkdir -p agents\u002Fhermes-life\ncp templates\u002Fagent\u002F*.md agents\u002Fhermes-life\u002F\n```\n\nThen fill in:\n\n```text\nagents\u002Fhermes-life\u002Finventory.md\nagents\u002Fhermes-life\u002Fdocker.md\nagents\u002Fhermes-life\u002Fenv-map.md\nagents\u002Fhermes-life\u002Frunbook.md\nagents\u002Fhermes-life\u002Fbackup.md\n```\n\nKeep raw secrets out of those files.\n\n### Option C: Bootstrap With The Setup Skill\n\nUse this if you have an SSH alias already configured locally.\n\nThe bundled `setup-control-room` skill is meant to:\n\n- connect to your VPS over SSH\n- install base packages\n- install Node.js\n- install Claude Code\n- install Codex CLI\n- install Docker\n- install Hermes Agent best-effort\n- clone this repo to `\u002Froot\u002Fagent-control-room`\n- link bundled skills into `~\u002F.claude\u002Fskills`\n\nAfter it runs, SSH into the VPS and finish interactive auth:\n\n```bash\nssh \u003Calias>\nclaude \u002Flogin\ncodex\nhermes\n```\n\nThen start using the Control Room:\n\n```bash\ncd \u002Froot\u002Fagent-control-room\ncat README.md\nls templates\u002Fagent\u002F\nls skills\u002F\n```\n\n### Recommended First Milestone\n\nDo not start by building a whole agent team.\n\nFirst milestone:\n\n```text\n1. Control Room exists on the VPS.\n2. One agent is documented in agents\u002F\u003Cagent-name>\u002F.\n3. No raw secrets are in the repo.\n4. You can restart\u002Fdebug\u002Frecover that one agent using its runbook.\n```\n\nThen move to Level 2 and add direct specialists.\n\n## Runtime Split\n\nKeep the control plane separate from live runtime state.\n\n```text\n\u002Froot\u002Fagent-control-room\n  docs, templates, runbooks, registry, architecture\n  no raw secrets\n\n\u002Fsrv\u002F\u003Cagent-name>\u002Fdata\n  live Hermes runtime\n  .env, memory, skills, sessions, crons, logs\n```\n\n## Security Rule\n\nNever commit raw secrets.\n\nThe control room may record:\n\n- secret names\n- provider\n- scope\n- location\n- rotation date\n\nIt must not record:\n\n- API key values\n- OAuth refresh tokens\n- passwords\n- private keys\n\n## License\n\nMIT. See `LICENSE`.\n","Hermes Agent Control Room 是一个用于管理和扩展 Hermes 代理的控制室模板。项目通过提供一套控制平面文件夹结构、运行手册模板、Docker 笔记、密钥映射及备份等，帮助用户从单一代理逐步构建至专业团队和自动化工作流。其核心功能包括层级架构设计以支持系统规模增长、任务总线模式实现调度器与专家代理间的委托作业以及集成的设置和操作技能，使代理能够自主构建或管理系统。适用于需要将多个 AI 代理协同工作的场景，如个人助理、SEO 优化、代码开发、市场营销等领域，尤其适合那些希望将 Hermes 代理作为操作系统而非孤立机器人来运行的开发者。",2,"2026-06-11 03:54:03","CREATED_QUERY"]