Agent Fleet

AI OS orchestrates 73 specialized agents powered by 6 AI models across 4 routing tiers. Claude Opus 5 is the default brain and powers the three reasoning tiers — Strategic (effort: xhigh), Professional (high), and Scout (low) — while the Specialized tier routes to the right model for the job: Gemini Omni for creative generation, DeepSeek for economy batch work, Grok (xAI) for real-time intelligence, OpenAI/GPT for alternate reasoning, and Perplexity for grounded search. Each agent has a defined role, capabilities, and escalation path managed through Markdown files with YAML frontmatter in .claude/agents/.

The fleet is 73 named employees across 11 departments on the Virtual HQ org chart, including the Knowledge & Records department. That figure is the whole fleet — the SEO Agency sub-agents, the AI Web Studio crew and the orchestration kernel each hold a seat on the chart rather than sitting outside it, so they are already counted here and are not a separate pool added on top.

Strategic Tier — Claude Opus 5 (effort: xhigh)

Maximum reasoning depth for decisions requiring deep analysis, architectural judgment, and cross-system awareness. Uses adaptive thinking for complex multi-step planning.

Agent	Role
Orchestrator	Central intelligence. Decomposes complex requests, delegates to sub-agents, manages inter-agent data flow, synthesizes final outputs.
Architect	System design and technical architecture. Evaluates trade-offs, defines data models, plans migrations.
Reviewer	Code and content review. Identifies quality issues, enforces standards, provides actionable improvement suggestions.
Security Auditor	Vulnerability assessment, dependency auditing, CSP review, authentication flow analysis.
Cost Analyst	Budget management, cost optimization across models and tiers, and financial reporting.
Growth Specialist (Summit)	Company-level growth strategy: the thesis, twelve-month goals with baselines from platform data, expansion options ranked, the plan to scale. Hands execution to marketing and the orchestrator.
Marketing Strategist (Vector)	Marketing department head: positioning, messaging, channel priorities, campaign plans and funnel experiments. Sequences the marketing executors; writes none of the copy itself.

Chief Librarian (Athena), the Knowledge & Records department head, is also Strategic tier — see Knowledge & Records below.

Professional Tier — Sonnet 5 (effort: high)

The workhorse tier handles the majority of tasks with balanced quality and speed. Under the default balanced reasoning mode it runs Claude Sonnet 5 at high effort, which is what keeps the bulk of agent work affordable; switching the mode to opus in Settings moves this tier onto Opus 5 instead.

Agent	Role
Researcher	Deep research with source synthesis, citation tracking, and structured output generation.
Coder	Full-stack development, debugging, refactoring, and test authoring across languages.
Writer	Long-form content, copywriting, documentation, and tone-adaptive writing.
Design System	DESIGN.md creation, brand reasoning, color theory, WCAG compliance, component specifications.
Media Producer	Remotion video generation, Google Vids integration, asset composition.
Lead Gen	Prospect scraping, enrichment, scoring, and personalized outreach sequence generation.
Marketing Hub	Multi-platform content pipelines, scheduling, A/B variant generation, performance tracking.
Product Factory	Digital product creation, marketplace listing, Etsy/Gumroad publishing automation.
Vibe Designer	Prompt-driven UI generation with predictive heat maps and interaction flow analysis.
Blender 3D	Text-to-3D environments and product renders via Blender MCP integration.
Synthesis	Cross-agent output merging, conflict resolution, and cohesive deliverable assembly.
Automator	CRON-scheduled routines, workflow triggers, and conditional automation chains.
Browser Agent	Headless browser control for scraping, form filling, and multi-step web interactions.
Report Compiler	Structured report generation from multiple data sources with charts and tables.
Research Architect	Research project planning, methodology design, and literature review structuring.
Data Wrangler	Data cleaning, transformation, format conversion, and statistical analysis.
QA	Test plan creation, regression testing, edge case identification, and quality metrics.
DevOps	Deployment, monitoring, infrastructure, and CI/CD pipelines.
Content Writer	SEO/AEO-aware page copy and structured content for generated and managed sites.
Factory	Agent and skill scaffolding — generates new agent definitions and pipeline templates.

Scout Tier — Sonnet 5 (effort: low)

Fast agents for quick lookups and lightweight tasks. Under the default balanced reasoning mode this tier runs Claude Sonnet 5 at low effort, which minimizes thinking time while retaining the model's full capability when a task needs it.

Agent	Role
Scout	Quick fact-checking, simple lookups, and rapid triage of incoming requests.
Social Intel	Social media monitoring, sentiment analysis, and trend detection.
Routine Runner	Executes pre-defined CRON routines with minimal overhead.

Specialized Tier — Economy (DeepSeek)

Part of the Specialized routing tier. Maximum throughput at minimal cost for batch operations.

Agent	Role
DeepSeek Worker	Bulk content generation, template expansion, and high-volume text processing.
Batch Runner	Queue-based mass production with automatic retry and output routing.

Specialized Tier — Realtime & Dev Planning (Grok / xAI) & Grounded Search (Perplexity)

Part of the Specialized routing tier. Grok handles live web intelligence; Grok Build plans platform upgrades; Perplexity Sonar provides grounded search with source citations and powers multi-model consensus in the AEO engine.

Agent	Role
Grok Realtime	Live data queries, current events, real-time web intelligence, and social platform analysis (Grok / xAI).
Dev Architect Grok	Plans upgrades to AI OS itself using xAI's Grok Build (grok-build-0.1), and proposes distribution blueprints as draft GitHub PRs — see Self-Improve. Read-only; every apply and every PR always requires human approval.

AI Web Studio Crew

Agents that design, build, and host static sites on the VPS — with custom domains, TLS, and import/export. They also keep generated sites passing AI OS's own AEO scorer.

Agent	Role
Web Studio Lead	Plans and orchestrates site builds — information architecture, design direction, and hosting strategy.
Web Builder	Generates the static site from the plan — HTML/CSS, JSON-LD, llms.txt, and robots.txt for AEO readiness.
Hosting Ops	VPS hosting, custom-domain binding, TLS provisioning, and publish/rollback of hosted sites.

Customer Service & IT

Client-facing support and internal infrastructure agents, available in Business and Enterprise tiers.

Agent	Role
CS Lead (Harbor)	Escalation management, ticket triage, and satisfaction tracking.
CS Tier 1 (Compass)	First-response support, FAQ handling, and basic troubleshooting.
CS Tier 2 (Resolve)	Complex issue resolution, technical investigation, and bug reproduction.
Customer Success (Keel)	Retention for the subscription base: onboarding health, renewal-payment failures, churn-risk reads, and win-back drafts. Drafts only; every message goes through the approval inbox.
IT Director (Matrix)	Infrastructure oversight, system health, and deployment coordination.
Sysadmin (Root)	Server management, monitoring, uptime, and security patches.
Help Desk (Guide)	Internal support, tool provisioning, and access management.
AI Helpdesk Concierge (Beacon)	Public-facing contact-page support agent — answers visitor questions from AI OS documentation, escalates unresolved issues to a logged ticket.

Agent Definitions

Each agent is defined as a Markdown file in .claude/agents/ with YAML frontmatter specifying its metadata:

---
name: researcher
model: opus-5
effort: high
tier: professional
escalates_to: orchestrator
tools:
  - web_search
  - vault_read
  - vault_write
---

# Researcher Agent

You are a deep research specialist. Your role is to...

Frontmatter Fields

• name — Unique identifier used in API calls and routing
• model — AI model to use; one of the 6 supported models (claude-opus-5, gemini-omni-flash, deepseek, grok, openai/gpt, perplexity)
• effort — Opus 5 effort level: xhigh (strategic), high (professional), low (scout)
• tier — Cost/capability tier for routing decisions
• escalates_to — Agent to escalate to when confidence is low
• tools — List of tools/capabilities available to this agent

Creative Tier — Gemini Omni

Multimodal creative generation powered by Gemini Omni Flash. Produces video, images, audio, thumbnails, and social clips from text prompts.

Agent	Role
Media Producer (Muse)	Creative pipeline orchestration, production strategy, brand consistency across media.
Vibe Designer (Pixel)	Prompt-driven UI generation, predictive heat maps, interaction flow analysis.
Video Creator (Reel)	Text/image/audio to video with physics simulation. Up to 1080p, 60s.
Audio Producer (Sonance)	Voiceovers, music, sound effects, podcast audio from text prompts.
Thumbnail Gen	Platform-optimized thumbnails with A/B variants for YouTube, social, blog.

SEO & AEO Agency Sub-Agents

Five specialized agents that run in parallel during SEO audits, each analyzing a different dimension of the target domain. The same fleet powers Answer-Engine-Optimization readiness — an 8-dimension score with multi-model consensus that measures how well a site is positioned for ChatGPT, Perplexity, and Google AI Overviews (readiness, not guaranteed citations or rankings).

Agent	Role
SEO Keyword	Keyword research, gap analysis, cannibalization detection, local targeting.
SEO Technical	Crawl analysis, Core Web Vitals, HTTP status codes, security, structured data.
SEO Competitor	Top 10 competitor profiling, content velocity, schema adoption, ranking overlap.
SEO Content	Content inventory, thin content detection, topic clusters, meta data quality.
SEO Backlink	Referring domains, toxicity detection, broken link recovery, anchor text analysis.

Legal Department

Handles licensing agreements, compliance, contracts, and intellectual property protection.

Agent	Role
General Counsel (Justice)	Chief Legal Officer. License agreements, IP protection, regulatory compliance, dispute resolution.
Compliance Officer (Shield)	GDPR/CCPA enforcement, audit trails, policy monitoring, API usage compliance.
Commercial Licensing Specialist (Covenant)	Software license agreements, software-licensing terms, SaaS licensing, usage rights.
Contract Specialist (Clause)	Contract generation, review, lifecycle management, template library.

Knowledge & Records

Department #11 — a governed catalog over the company's existing document stores (the Memory Vault, extracted uploads, and agent-output artifacts), not a fourth store. Every agent on every tier reads it, so library content only ever reaches an agent inside the same fenced, untrusted-by-default envelope used for outside data.

Agent	Role
Chief Librarian (Athena)	Department head. Taxonomy authority, cross-department lookup, routing, and retention decisions (legal hold is a consult with the Legal department, not this agent's call). Reads only, never ingests.
Archivist (Vellum)	Intake, format handling, dedupe, metadata, and versioning — extends the existing document-extraction pipeline rather than reimplementing it.
Knowledge Manager (Archive)	Auto-categorizing knowledge ingestion, semantic linking, and radial graph visualization.
Sync Steward (Tether)	Source-change detection and knowledge-base re-sync — Gemini Gems synced to NotebookLM notebooks for continuous knowledge amplification.

System & Orchestration Kernel

The Dynamic-Workflow orchestration kernel coordinates the whole fleet. Most kernel roles are filled by agents already listed above — Synthesis (output merging), Sync Steward (knowledge amplification), Report Compiler, Routine Runner, Batch Runner, Factory, and Content Writer — plus the dedicated Safety agent below.

Agent	Role
Safety	Auto-Mode approval gating — classifies irreversible/outward-facing actions and enforces human-in-the-loop approval before they run.

Persistent Layer — Hermes MCP

The always-on background worker layer. Hermes runs 24/7 on the platform and handles tasks that require persistence beyond the active session. These are operating modes of the platform's persistent worker (mapped to the Operations department's Hermes desks) rather than additional fleet agents.

Agent	Role
Hermes Delegate	Background task execution with progress pings. Walkaway mode for long refactoring tasks with mobile replies.
Hermes Cron	24/7 scheduled jobs: inbox summaries, news briefs, GitHub backups, comment monitoring, health checks.
Hermes Approval Gate	Pauses destructive actions and pushes approval prompts to Telegram/Slack/Discord for mobile review.

LLM Provider Consultants

Seven on-site LLM provider consultants advise the operator on model releases and how to adopt them inside AI OS — one each for Anthropic, OpenAI, Gemini, DeepSeek, Grok, Perplexity, and Manus. Each consultant answers on its own provider’s model (the OpenAI consultant on GPT-5.6, the Gemini consultant on Gemini 3.5 Flash, Perplexity’s on Sonar Pro with citations, and so on), so it is genuinely grounded in the stack it advises on. Manus has no API caller on the platform, so its consultant runs on Claude with a Manus knowledge pack and says so plainly.

Each carries a verified knowledge pack (current model IDs, pricing, and breaking changes) plus live web search, and is disciplined about re-verifying anything time-sensitive before presenting it. Ask one for the latest from its provider, or for concrete guidance on wiring a new release into the platform’s routing and callers.

Communications Director

The Communications Director (Herald) is the single channel through which the platform’s model and architecture intelligence reaches its audiences. When the Orchestrator or Architect makes a model/provider decision — consulting the LLM provider consultants along the way — the decision and the consultants’ findings flow to the Communications Director, who synthesizes across the sources, reconciles their confidence levels, and produces the audience-appropriate artifact: an operator briefing, a change/release note, a team announcement, or an external draft (publishing is outward-facing and routes through the human-approval gate). It makes decisions understood; it does not make them (that stays with the Orchestrator and Architect) and it never publishes on its own.

Escalation Paths

Agents can escalate tasks upward when they encounter requests beyond their capability. The escalation chain follows the tier hierarchy:

• Specialized / Scout → Professional → Strategic (Orchestrator)
• The Orchestrator is the terminal escalation point and can re-delegate to any agent

Frequently asked questions