↑

// SERVICE SPECIFICATION 03

Claude AI Agent Orchestration 101

The principles behind Claude AI agent orchestration and how to run agent teams at scale. Build multi-agent workflows that automate complex operations while keeping quality and cost under control, even on a small budget.

Commission AI Agent Swarm Orchestration →

1. Service Overview & Scope

  • Claude Agent Teams: Coordinating several Claude agents on one goal, with an orchestrator on Opus or Fable at high or xhigh effort and Sonnet or Haiku workers at lower effort.
  • Five Core Architectures: Orchestrator-worker, hierarchical, sequential pipeline, parallel fan-out and fan-in, and router, with guidance on when each one fits.
  • Model Context Protocol (MCP) Integration: Connecting agents to internal APIs, databases and files through MCP servers with defined input schemas.
  • Verification Before Commit: Reviewer subagents and test suites check worker output before it changes anything.
  • Spend Limits at Every Level: Budgets per session, day, month and year, iteration caps, and a kill switch that stops concurrent cloud agents.

2. Standard Architecture & Tools

  • Orchestration Layer: Claude Code subagents, background agents in their own git worktrees, and the Claude API with Python asyncio for concurrent calls.
  • Model Tier Allocation: Claude Opus or Fable (planning and final synthesis), Claude Sonnet (coding, tool use and analysis), and Claude Haiku (search, classification and summarization).
  • Shared State: Task and result state kept in files or a SQL database, so interrupted runs can resume.
  • Cost and Safety Controls: Prompt caching, Batch API fan-out, a restricted tool list for each subagent, Docker sandboxes, and run logs that record tokens per agent.

3. Deliverable Examples & 4-Week Schedule

  • Orchestration Repository: Documented Python codebase with the orchestrator, worker definitions and routing logic, plus its tests.
  • Custom MCP Server Suite: MCP servers in Python or Rust connecting your agents to SQL databases, ticket queues and internal APIs.
  • Checkpoint & Resume Logic: Saved state that lets a multi-hour agent run pick up where it stopped.
  • Cost & Run Report: Per-agent breakdown of tokens, cost, tool calls and failures for each run, with a worked budget for your workload.
Timeline Focus & Architecture Core Deliverables & Artifacts Verification & Sign-off Criteria
Week 1 Architecture Selection & Agent Topology Design Architecture specification selecting from the five core patterns (orchestrator-worker, hierarchical, sequential pipeline, parallel fan-out/fan-in, router), model tier assignments (Opus or Fable lead, Sonnet or Haiku workers), and effort level tuning Stakeholder sign-off on agent team topology, effort budgets, and git worktree isolation model
Week 2 MCP Tooling & Async Orchestration Engine Custom Python or Rust MCP servers with defined input schemas for SQL, ticket queues, and internal APIs, Python asyncio orchestration engine, and Docker sandbox setup Verified schema-enforced tool execution between subagents and MCP servers within restricted container boundaries
Week 3 Verification Graphs, State Checkpoints & Spend Controls Reviewer subagent verification logic, automated test suites before commit, SQL or file state checkpoint & resume engine, prompt caching, and per-session/daily token budgets Verified state recovery after simulated run interruption and reviewer subagents catching errors before commit
Week 4 Batch API Fan-out, Cost Reporting & Handover Batch API processing pipeline, per-agent token and cost telemetry reporting, hard kill-switch breakers, complete codebase transfer, and 30-day SLA onboarding Multi-agent stress run completed with zero runaway loops, confirmed budget ceiling enforcement, full IP transfer, and client sign-off

4. Guarantees & Commitments

  • No Runaway Loops: Iteration counters, timeouts and token caps on every agent.
  • Full IP & Code Ownership: All orchestration code, prompts, skills and MCP configurations transfer to you.
  • Enterprise Data Boundaries: Agents act with each user's own permissions, and write actions need approval until you choose to relax them.
  • 30-Day Production SLA: Support channel for adjustments, new tools and debugging.

5. Investment & Pricing Quote

  • Modular Service Selection & Flexible Engagement: Clients can freely pick and choose their services, either individually or bundled under our comprehensive transformation programs:
    • Individual Service Pick (Personal / 1-on-1): $100 USD per service for 5 days of live training + 5 days of dedicated support (10 days total coverage per service). Strictly 1 person only.
    • Individual Service Pick (Business / Teams): $200 USD per service for 5 days of live training + 5 days of dedicated support (10 days total coverage per service). Covers up to 20 participants (additional seats at $100 USD per seat).
    • Comprehensive Business Training: $2,000 USD for 60 days (30 days live training + 30 days guidance and detailed guides and materials). Includes cohorts up to 20 participants. Extra seats at $100 USD per seat).
    • Comprehensive Personal Training: $1,000 USD for 30 days (15 days live training + 15 days guidance and detailed guides and materials). Strictly for 1 person only (private acceleration).
  • Global Purchasing Power Parity (PPP) Available for All Service Pricing:
    • Tier 1 (Standard / 100%): Individual $100 USD | Business $200 USD | Business 60d $2,000 USD | Personal 30d $1,000 USD (USA, UK, EU, Canada, Australia)
    • Tier 2 (Developing Nations / 25% Off): Individual $75 USD | Business $150 USD | Business 60d $1,500 USD | Personal 30d $750 USD (India, Brazil, Mexico, UAE, Poland)
    • Tier 3 & 4 (Lower-Middle/Low-Income & Startups / 50% Off): Individual $50 USD | Business $100 USD | Business 60d $1,000 USD | Personal 30d $500 USD
    • Tier 5 (Academia & Non-Profits / 75% Off): Individual $25 USD | Business $50 USD | Business 60d $500 USD | Personal 30d $250 USD
    • Tier 6 (Crisis-Affected & Under-Resourced Regions): Individual $10 USD | Business $20 USD | Custom Subsidized Relief / Micro-Rates
    • Calculate and lock in your exact tier rate live on the Pricing page.
  • Payment Schedule:
    • Zero upfront advance; billed in 5 equal milestone installments due only after verified instruction and runtime delivery.

6. Summary

  • Claude Agent Teams: Opus or Fable planning paired with Sonnet and Haiku workers.
  • Strict 1-Month Delivery: 4-week progression from workflow breakdown to production handover.
  • Model Context Protocol Tools: MCP servers connecting agents to SQL, internal APIs and file systems.
  • Low-Budget Operation: Prompt caching, batch processing, cheaper models for narrow jobs and hard spend limits.
  • Complete Code & IP Transfer: Full ownership, containerized runtimes and 30-day support.
Commission AI Agent Swarms → Book a free intro call on LinkedIn ↗