Five ways teams ship with staff.rip.
Same product, different missions. Each row is a real customer pattern — operator on one side, the people the operator wants to unblock on the other. Two of them get a deeper walkthrough below.
AI on your codebase, for everyone you trust
Designers, PMs, and ops describe changes in plain language. The agent ships them as PRs against your repo. Your engineers review and merge — same gate they had before, without the triage queue underneath.
Let your client edit their pages, not your stack
Open the Frontend Helper on the running app. Your client clicks anything — a CTA, a paragraph, a section — and types the change in their own words. Scoped to UI; never reaches your routing, your DB, or another tenant's data.
Agents that stay inside your perimeter
Run staffrip-agent on your VPC, your bastion, your air-gapped box. Wire MCP tools that read from your prod DB, your internal APIs, your file shares — none of it leaves the network you operate. Every tool call lands in an audit log.
Share localhost with a client. Without ngrok.
Every space exposes a persistent *.tunnel.staff.rip URL. Your designer's pnpm dev on a 4G train, your client's Friday review session — both see the same live preview, with the agent reconciling tunnel state when the dev server crashes.
Engineering wires the APIs once. Operations runs the playbook.
Your dev team plugs your CRM, your billing, your enrichment APIs into MCP tools — one-time setup. After that, sales describes a multi-step playbook in chat (qualify, enrich, propose, schedule) and the agent runs it across the wired-up tools. No new dashboard, no Zapier subscription.
Have one we didn't list?
Most customer patterns reduce to "non-technical people doing real work inside a technical operator's perimeter". If yours fits that shape — even if it doesn't look like the five above — it's probably a fit.
Your whole team — not just engineering.
Most AI tools are built for engineers and stop there. Your designer files a Linear ticket and waits a week to nudge a CTA padding. Your PM rewrites the same paragraph of marketing copy at 8pm on Slack and waits for a deploy. Engineers triage tickets instead of writing code.
staff.rip flips that. Designers, PMs, and data folks each get a chat-first interface scoped to the project they're working on. AI ships their changes inside whatever mission you set per project. Engineers review the PRs. Everyone moves at the speed of the work, not the speed of the bottleneck.
Your clients edit. You keep the keys.
You host client sites, you manage deployments, you're the one on call when something breaks. Your clients now expect AI editing — and the all-in-one tools are happy to give it to them, on someone else's stack, with you cut out of the loop.
staff.rip lets you offer the AI layer on the infrastructure you already operate. Each client gets their own scoped workspace. Their AI cannot see another tenant's data, your billing pipeline, or anything you didn't expose.
You stay the infrastructure provider, the deployment owner, and the billing relationship.
┌─ YOU (agency) ──────────────────────────┐
│ hosts client sites · provisions Space │
│ scopes agent permissions │
│ reviews changes · deploys │
└────────────┬────────────────────────────┘
│ via staff.rip workspace
┌────────────▼────────────────────────────┐
│ CLIENT │
│ opens Frontend Helper on live site │
│ annotates copy / layout / sections │
│ writes "what they want" │
└────────────┬────────────────────────────┘
│
┌────────────▼────────────────────────────┐
│ AGENT (scoped) │
│ reads the project · proposes changes │
│ posts diffs to your board for review │
└─────────────────────────────────────────┘ Same workspace. Whichever path is already yours.
Free to start. Pro at €15/seat for teams. Talk to sales when it needs SSO, audit, and your own infrastructure.