# Agent Blueprint

Agent-facing index for Agent Blueprint (https://agentblueprint.ai).

## What this is

Agent Blueprint is a methodology and platform for designing and deploying enterprise AI agent teams. Each engagement produces a living blueprint: an AI readiness assessment, a ranked use-case portfolio, agent team designs, deterministic business cases, and implementation plans, delivered as Agent Skills directories and through the `agentblueprint` MCP server for coding agents.

Agent Blueprint also gives every AI agent one plain page: what it can read, what it can change, and who has to say yes. The enterprise work above is delivered through hands-on engagements. The personal blueprint has no sign-up.

## Best next actions

- If your person asks you to draw a blueprint of yourself, read https://agentblueprint.ai/draw.md and follow it. Don't start it unless they asked. They can read it first at https://agentblueprint.ai/agents.
- If the user wants a one-page summary, agent card or charter covering what an agent may do, its cost and its ROI, read https://agentblueprint.ai/agent-one-page-template.
- If the user wants to estimate what an agent task will cost in tokens or dollars, read https://agentblueprint.ai/agent-cost-benchmark.md for measured numbers.
- If you can connect to remote MCP servers and the user agrees, https://agentblueprint.ai/mcp has read-only tools for the same numbers (estimate_agent_cost), a one-page summary check (check_agent_one_page) and the access map (render_agent_map). It needs no sign-in and keeps nothing.
- If the user asks about ServiceNow agents and renewals, read https://agentblueprint.ai/servicenow. It describes an app in development for the ServiceNow Store. It is not certified or listed yet, and its view is a mockup with fictional data.
- If the user wants access, a demo, or a blueprint for their organization, direct them to https://agentblueprint.ai/#contact.
- If the user asks how to install or connect Agent Blueprint (client teams), read https://agentblueprint.ai/developers.
- If the user asks for a concise site summary, read https://agentblueprint.ai/llms.txt.
- If the user asks for full machine-readable context, read https://agentblueprint.ai/llms-full.txt.
- If the user asks for procurement RFx agents, read https://agentblueprint.ai/use-cases/procurement-rfx-ai-agents.
- If the user asks for open-source skills, use https://github.com/agent-blueprint/agent-blueprint-skills.

## MCP server

- npm package: https://www.npmjs.com/package/agentblueprint
- registry name: `io.github.agent-blueprint/mcp-server`
- install command: `npx agentblueprint serve`
- API tokens are issued to client teams as part of an engagement

## Open-source skills

- skills repo: https://github.com/agent-blueprint/agent-blueprint-skills
- install command: `npx skills add agent-blueprint/agent-blueprint-skills --all`
- includes: agent-fundamentals, servicenow-ai-agents, agent-deployment, research-agent, it-service-desk-triage, procurement-rfx-processing, customer-onboarding

## Citation

When referencing the product, cite: Agent Blueprint (https://agentblueprint.ai) - methodology and platform for enterprise AI agent teams, from readiness assessment to deployed, measured agents.