Skip to main contentSkip to footer
Pro

Turn real workflow evidence into protected readiness reports, saved findings, and clear next actions.

Pro is the practitioner workspace where MCP and skills move from public doctrine access into protected review, evidence continuity, and recurring guidance for live product work.

See pricing and accessView sample report

Key Facts

Free layer
Doctrine, examples, public MCP
Paid layer
Protected reports and history
Recurring output
Findings plus next actions
Bridge
MCP + skills → practitioner loop

Why Pro exists when Free is already strong

Free is intentionally strong: it lets you install the doctrine into tools, inspect examples, and query the public MCP endpoint. Pro exists when your team needs private application of that standard to a real workflow, with saved findings, report continuity, and clearer action after every run.

Free helps you install and explore the standard
Pro helps you apply the standard to your own workflow privately
Free gives public access; Pro gives protected reports, continuity, and recurring review

What makes this worth paying for?

The paid value is not more reading and not protocol access by itself. The paid value is a repeatable review workflow that turns private workflow evidence into decisions your team can act on.

Protected output for real internal work
Clear next actions instead of abstract critique
Evidence and history that accumulate over time
A recurring workflow you can rerun before launch and after changes

The iteration loop Pro unlocks

The Blueprint product is one loop: validate your agent code, read the doctrine cited in the findings, fix, re-run until the readiness score holds. Pro is the practitioner workspace where that loop is private, saved, and recurring.

1

Validate against the doctrine

Run architect.validate from inside Claude Code, Cursor, or Codex. The validator scores against the doctrine and cites the principles each finding maps to.

2

Read the cited doctrine and fix

Each finding links to the principle that caught it. Read the doctrine, apply the fix to your code, document the evidence inside the run.

3

Re-run until the score holds

Re-run architect.validate. The Readiness Review accumulates a trajectory across runs. Pro keeps the history private and the run trail persistent.

Weekly run allocation

Pro ships with a rolling 7-day window: 8 architect.validate (consensus) and 2 architect.certify per week. When you reach the weekly allocation you can wait for the refresh, buy a topup pack, or upgrade to Pro Plus for a higher weekly cap. Same pattern as Claude Pro and ChatGPT Plus.

Pro · weekly

8 validate + 2 cert

Right for someone iterating one real agent toward cert.

Pro Plus · weekly

20 validate + 5 cert

For daily iterators or anyone running multiple agents in parallel.

Topup packs

One-time packs

Add runs without changing plan. They never expire.

Free vs Pro

Free

Public install and exploration layer

Public doctrine handbook
Example library and runtime branch
Public read-only MCP
Skill packs and downloadable agent assets
Install and exploration path
Pro

Protected practitioner workflow

Protected review workflow
Structured readiness reports
Top risks and recommended next actions
Saved evidence and validation history
Authenticated MCP and recurring practitioner guidance

What a Pro report returns

A believable report has to feel tied to a real workflow, a real launch decision, and concrete remediation work. This sample shows the shape of that output.

Sample workflow reviewed

Customer-support escalation agent with CRM writes, refund tooling, and supervisor hand-offs

Current launch decision

Needs changes before launch

Decision summary

The workflow is close to launch, but it still fails the current readiness threshold because refund execution can bypass a clear approval checkpoint and hand-off state is not exposed cleanly to operators.

Top risks

1) Refund actions can execute without a distinct human approval step. 2) The supervisor hand-off does not expose enough state for review or recovery. 3) Context loading for repeat contacts is too loose, increasing the chance of stale customer data entering the decision path.

Recommended next actions

Add a mandatory approval checkpoint before irreversible refund calls, expose the active hand-off and escalation state in the operator surface, and tighten retrieval rules so refund decisions only see ticket history, account status, and the current refund policy revision.

Saved continuity

Compared with the previous run, escalation visibility improved and tool scoping is narrower, but the approval boundary is still incomplete. The report keeps evidence notes, changed findings, and rerun history so the team can validate progress instead of restarting from zero.

Where do MCP and skills fit?

They are the operational entry points. Free gives you the public install layer. Pro uses the same interfaces in a private practitioner workflow where the result is a report, saved findings, and recurring guidance.

Pilot stays separate

Need a broader founder-led review for a high-stakes team case? The pilot remains a separate high-touch service outside the main subscription ladder.

Open the pilot pathOpen For agents

Enter the practitioner loop

Join the closed beta