MCP setup
Connect your agent runtime to the public Blueprint MCP endpoint in three steps.
https://aidesignblueprint.com/mcp- Streamable HTTP
- Protocol 2025-11-25
- Public tools need no key
Set up in three steps
Connect your client
Add the server to your agent once. Each tab has the exact command or config; the public tools work without a key.
Terminal
claude mcp add aidesignblueprint --transport http https://aidesignblueprint.com/mcpRun in your terminal, then use /mcp inside Claude Code to sign in. No token needed: the OAuth prompt covers the paid validators.
.cursor/mcp.json
{ "mcpServers": { "aidesignblueprint": { "type": "http", "url": "https://aidesignblueprint.com/mcp" } } }The button opens Cursor and installs the server. Or save the config at the path above, then restart Cursor.
.vscode/mcp.json (server entry)
{ "name": "aidesignblueprint", "type": "http", "url": "https://aidesignblueprint.com/mcp" }The button opens VS Code's install prompt. Or add the server entry to .vscode/mcp.json.
Terminal
codex mcp add aidesignblueprint --url https://aidesignblueprint.com/mcpRun it in your terminal, then start Codex.
Terminal
gemini mcp add --transport http aidesignblueprint https://aidesignblueprint.com/mcpRun in your terminal, then restart your Gemini CLI session.
Add this entry to your client's MCP config
MCP config
{ "mcpServers": { "aidesignblueprint": { "type": "http", "url": "https://aidesignblueprint.com/mcp" } } }Any client that speaks Streamable HTTP can use it. The public tools need no key.
To use the validators, sign in: for clients other than Claude Code, create a token on Connect your coding agent.
Not on MCP: Windsurf, GitHub Copilot, DeepSeek and Qwen read the doctrine from files; the file for each is on Start here.
Check the session
Four calls from a terminal confirm the transport, the session and a real tool before you rely on the agent. Paste them in order: the first keeps the session id in SID for the others.
initializenotifications/initializedtools/callDELETE
Open a session
initializereturns the session id in themcp-session-idheader; the command keeps it inSID.Terminal
SID=$(curl -s -D - -o /dev/null -X POST "https://aidesignblueprint.com/mcp" \ -H "Accept: application/json, text/event-stream" \ -H "Content-Type: application/json" \ -d '{ "jsonrpc": "2.0", "id": 1, "method": "initialize", "params": { "protocolVersion": "2025-11-25", "capabilities": {}, "clientInfo": { "name": "manual-check", "version": "1.0.0" } } }' | awk 'tolower($1)=="mcp-session-id:" {print $2}' | tr -d '\r') echo "$SID"Expected
An id of 32 hexadecimal characters. An empty line means the session did not open.
Confirm the session
The
notifications/initializedmessage tells the server the client is ready.Terminal
curl -s -o /dev/null -w "%{http_code}\n" -X POST "https://aidesignblueprint.com/mcp" \ -H "Accept: application/json, text/event-stream" \ -H "Content-Type: application/json" \ -H "Mcp-Session-Id: $SID" \ -d '{"jsonrpc": "2.0", "method": "notifications/initialized"}'Expected
202Call a tool
clusters.listis the smallest real call: if it answers, the server is reachable and the public tools respond.Terminal
curl -s -X POST "https://aidesignblueprint.com/mcp" \ -H "Accept: application/json, text/event-stream" \ -H "Content-Type: application/json" \ -H "Mcp-Session-Id: $SID" \ -d '{ "jsonrpc": "2.0", "id": 2, "method": "tools/call", "params": { "name": "clusters.list", "arguments": {} } }'Expected
Four clusters: delegation, visibility, trust and orchestration.
{"jsonrpc":"2.0","id":2,"result":{…Close the session
DELETEwith the same id ends the session on the server.Terminal
curl -s -o /dev/null -w "%{http_code}\n" -X DELETE "https://aidesignblueprint.com/mcp" \ -H "Mcp-Session-Id: $SID"Expected
200
Make your first call
With the server connected, start from a real task: paste one of these prompts into your agent.
Download the prompt packArchitecture audit Use the blueprint as an audit framework. List the clusters first, then propose which principles to use to assess this agent architecture and which examples to read next.
Example lookup Search examples for orchestration, visibility, and steering. Group them by principle and tell me which ones are worth reading first.
Principle explainer Explain the most relevant principle for this workflow with its definition, rationale, risk, and one linked example.
Tool reference
29 tools in four tiers. Arguments in grey, marked "?", are optional.
principles.get(slug) and examples.get(slug) are faster and cost your model fewer tokens; when you do not, use principles.search or examples.search. Keep the list calls for browsing.No key and no sign-in.
clusters.list()
See the four clusters and the principles in each. A good first call.
principles.list(cluster?, lens?)
Browse every principle, or one cluster's, with their slugs.
principles.get(slug, lens?)
You know the principle's slug and want its full text.
principles.search(query, limit?, lens?)
You can describe the problem but cannot name the principle.
clusters.get(slug)
You know a cluster's slug and want its principles.
examples.search(query, principle_ids?, difficulty?, library?, limit?)
Find code examples by topic, principle, difficulty or library.
examples.get(slug)
You know an example's slug and want the whole example.
guides.search(query, limit?)
Find the application guide for an engineering problem.
guides.list()
Browse the application guides.
guides.get(slug)
Read one application guide in full.
assets.list()
Get the download links for the doctrine files and agent packs.
signals.feedback(task_type?, surface?, rating_clarity?, rating_usefulness?, …)
Send feedback, only when the user asks you to.
Any signed-in account, from Basic up.
me.learning_path()
Read your learning path: course progress and certification status.
me.coaching_context()
See which stages still have thin evidence.
me.add_evidence(course_slug, stage_id, note)
Add an evidence note to a stage of your active course.
handoffs.operator(reason, topic?, …)
Hand off to a person: coaching for a builder working alone.
handoffs.partnership(reason, topic?, organization?, …)
Start a partnership conversation: co-building, distribution or collaboration.
handoffs.agency(reason, company?, …)
Ask the Blueprint agency for hands-on delivery support for your team.
A Pro, Pro Plus, Teams, Enterprise, beta or trial plan. architect.validate, design.validate, and spec.validate accept private_session=true to skip the stored run (operational security and cost logs are kept per the Privacy Policy).
architect.validate(implementation_context, focus_area?, task?, language?, repository?, files?, goals?, example_limit?, private_session?)
Review agent code against the 10 principles. Pass the same repository on every round so each run builds on the last.
design.validate(implementation_context, task?, goals?, files?, repository?, private_session?, session_id?)
Review a UI surface against the 8 experience design laws.
spec.validate(implementation_context, task?, goals?, files?, repository?, private_session?, session_id?)
Review a written spec against the 8 spec quality laws before anyone builds from it.
architect.validate_consensus(implementation_context, n?, focus_area?, …)
Run the review several times to see how stable the grade is.
architect.certify(run_id, code?)
Certify a standard run that reached production_ready, with a deeper pass.
me.validation_history(run_id?, repository?, limit?)
See your past runs, or fetch a run whose call timed out on your side.
me.sessions(session_id?)
List your Governed Sessions, or open one.
me.session_event(session_id, event_type, actor, summary)
Post a team event, such as a handoff, a steer or a gate, to a Governed Session that has team agents turned on.
me.await_steer(session_id, after_event_id?, timeout_s?)
Wait for the next steer from AIDB Studio in a team session.
signals.report(event_type, surface_used?, perceived_value?, …)
Record that a run was useful, when the user agrees.
A Teams or Enterprise plan.
team.summarize(days_back?, private_session?)
Summarise your team's tool usage and value signals from the logs.
Long runs as tasks
architect.validate can run as a task: you get an id at once and read the result when it is ready, with tasks/get and tasks/result.
Clients that declare the tasks capability add task to the params of an architect.validate call (signed in, with a plan that includes the validators). The server answers at once with the task, runs the review in the background and sends its status on notifications/tasks/status.
tools/call
{
"jsonrpc": "2.0",
"id": 3,
"method": "tools/call",
"params": {
"name": "architect.validate",
"arguments": { "implementation_context": "<your code>" },
"task": { "ttl": 600000 }
}
}| Method | What it does |
|---|---|
| tasks/get | The task's state and how often to poll |
| tasks/result | The result, once the task has finished |
| tasks/list | Your tasks, one page at a time |
| tasks/cancel | Stops a running task |
| notifications/tasks/status | Status updates the server pushes |
Only architect.validate runs as a task today; architect.validate_consensus and architect.certify answer within the call. The task id is the run id, so me.validation_history(run_id=…) reads the same result, which is also how you recover a call that timed out on your side. Standard runs only: a private session answers within the call and keeps no run to recover.
Also listed on
The directories that list this MCP server (com.aidesignblueprint/blueprint) and the Python package. More appear as they go live.
Also in this section