Launch Audits.

Stack Audit

You bought the seats. Quality still depends on who typed. I read the stack - Cursor, Claude, agents, CI, the review loop, and whether an agent can sign, spend, delete, deploy, or send without a person - and send back a written diagnosis.

You get

• How the stack is used today
• What is working, what is theater, what to change first
• An eval plan for up to three of the rules and gates you rely on most, so you can tell which ones change agent behavior
• Whether an agent can sign, spend, delete, deploy, or send without a person
• Whether you can prove what an agent did from a trail you can join
• Whether skills, tools, and MCP capabilities are inventoried, and whether pre-use review is something the skill under review cannot talk its way past
• Optional Loom ≤20 min if you attach a read-only repo or recording

Not included

hands-on install, retainer, building the product (Factory Pack).

Launch Hardening Audit

You shipped a vibe-coded SaaS. I read auth, secrets/spend, Stripe, tenancy, deploy, and the launch gate - and send a scorecard plus a 30-min readout before you put real traffic on it.

You get

• Written scorecard across auth, secrets/spend, Stripe, tenancy, deploy, launch gate
• Prioritized go / no-go notes
• 30-min readout (async scheduling after delivery)

Not included

hands-on install, CIAM/monitoring product, building the product (Factory Pack / Web+Stripe MVP).

Bundle

Both for $1,200 (save $300). Same 5-business-day async clock once payment + intake land.

Free setup review

Send me your CLAUDE.md, hooks, and settings. I'll send back what I think the setup is trying to do, and the first 3 things I'd change.

You send

CLAUDE.md, hooks, and settings. Strip any keys or tokens first.

You get

A short written read of what the setup is for, the top 3 fixes ranked, and where an eval would show whether a rule works. The full running stack and the eval plan itself are the Stack Audit.