← Ship an Agent ProductArrows or space · F for fullscreen · Print to saveTalks

SHIP AI TALKS

Ship an
agent product

Leave with an agent spec, a deployed support agent and the limits that let you charge for it.

Ship AI · TalksShip an Agent Product01/21

Who's hosting

How you do anything is how you do everything.

Santos Hernandez
Santos HernandezFounder, Ship AI
$25k+/moggbucks, built with agents
$0 to $12MARR at ZBD
EU firstMiCAR approval
THE COMMUNITY

Ship AI

Free workshops and talks in Phoenix.

shipai.club
THE STUDIO

Ship AI Studio

AI built into your business, in two weeks.

shipaistudio.com

Ship AI · TalksShip an Agent Product02/21

Join the room

SHIP AI DISCORD

Scan it now.
Prompts later.

Open the Discord invite

Ship AI · TalksShip an Agent Product03/21

Our partner

Ship AIWorkuityAI Exchange

Building Phoenix's AI community, together. Workuity hosts tonight, and we present Founder Weekend side by side.

Ship AI · TalksShip an Agent Product04/21

THE IDEA

A demo answers.
A product knows its limits.

Ship AI · TalksShip an Agent Product05/21

Demo or product

A DEMO

One prompt. Works on stage.

A PRODUCT

Tools, approvals, limits.
Evals before customers.

Ship AI · TalksShip an Agent Product06/21

Tonight

01

Spec

02

Tools

03

Approve

04

Guard

Ship AI · TalksShip an Agent Product07/21

Tonight's agent

AGENT-SPEC.md
Job

Answer a candle shop's order and return questions.

Tools

Look up orders, track, read policies, hand off.

Approval

Refunds wait for the owner's yes in Slack.

Done when

Wrong email reveals nothing

Approve refunds once

13 evals pass

Ship AI · TalksShip an Agent Product08/21

Vercel's agent stack

01

Chat SDK

Web chat and Slack, one bot

02

AI SDK

Agent, tools, streaming

03

Workflow SDK

Refunds that wait for a yes

04

Vercel

Deploy, env vars, logs, spend

Ship AI · TalksShip an Agent Product09/21

Spec the agent

claude · ~/deskhand

claudeWhich tools can move money?

youOnly refunds. The owner approves those.

claudeWhat may we log?

✓ AGENT-SPEC.md: 5 tools, 1 approval, 13 evals

Copy · write the agent spec
Interview me about Deskhand, then write AGENT-SPEC.md: job, users, tools, approvals, never list, guardrails, evals. Wait for my OK.

Full prompt

Copy this entire prompt into Claude Code.

I'm building Deskhand, a support agent for small online stores, starting with one Phoenix candle shop. Shoppers chat on the store's site. The owner approves anything that moves money from Slack.

Interview me one question at a time and wait for each answer: the job, who the users are, which tools the agent may call and what each one reads or writes, what must wait for the owner's approval, what the agent must never do, which limits protect us (rate, cost, message length), what we may and may not log, and how we'll know it works.

Then write AGENT-SPEC.md with these sections: Job, Users, Stack, Tools (a table: tool, reads or writes, approval), Never, Refund workflow (numbered steps), Guardrails, Evals, Done when. The stack is Chat SDK with the Web and Slack adapters, the AI SDK through Vercel AI Gateway, the Workflow SDK, and Vercel.

Don't write code yet. Stop and wait for my OK.

Ship AI · TalksShip an Agent Product10/21

Scaffold with Chat SDK

~/deskhand
ONE BOT, TWO CHANNELSnpm create chat-sdk@latest -- deskhand
--adapter web slack memory -y

Webhook routes, bot.ts and .env.example.

01

You add the skills

Chat SDK and Workflow SDK, from their own docs.

02

Claude scaffolds

The CLI runs without prompts inside Claude Code.

03

One folder

Move the spec in, restart Claude Code there.

Copy · scaffold the bot
I add the Chat SDK and Workflow SDK skills. Then scaffold with create-chat-sdk (web, slack, memory), move the spec in, and restart.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md.

1. In my own terminal I'll run:
npx skills add vercel/chat
npx skills add https://github.com/vercel/workflow --skill workflow
Wait until I tell you it's done.
2. From this parent folder (where AGENT-SPEC.md is, not inside a deskhand folder), run: npm create chat-sdk@latest -- deskhand --adapter web slack memory -y
Memory state is for local development only. We switch to Redis before deploying.
3. Move AGENT-SPEC.md into the deskhand folder. Show me the generated files and what each one does.
4. Tell me to restart Claude Code inside the deskhand folder so the skills load there. Stop and wait.

chat-sdk.dev

Ship AI · TalksShip an Agent Product11/21

The storefront chat

claude · ~/deskhand

✓ proxy.ts · signed visitor cookie

✓ app/page.tsx · Chat pattern on useChat

✓ onDirectMessage · streaming fullStream

claudeRun npm run dev and ask about returns.

Copy · the chat UI
A signed visitor cookie from proxy.ts, the shadcn/ui Chat pattern on useChat, and a streaming ToolLoopAgent through AI Gateway.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md, src/lib/bot.ts and src/app/api/chat/route.ts. Use the Chat SDK skill and the docs in node_modules/chat/docs.

1. Issue the visitor cookie in src/proxy.ts. Next.js 16 renamed middleware to proxy, so check https://nextjs.org/docs/app/api-reference/file-conventions/proxy first. When a request has no valid cookie, create a random visitor id (no colon in it), sign it with COOKIE_SECRET, and set it on the response as httpOnly, secure and sameSite lax. Skip static files in the matcher.
2. Replace the getUser stub so it only reads and verifies that cookie: return { id } when the signature checks out, and null (a 401) when the cookie is missing or invalid. getUser never creates a cookie. Tell me to add COOKIE_SECRET to .env.local myself (a long random string) for local dev, and wait until I tell you it's done.
3. The scaffold has no pages, no root layout and no Tailwind. Add a root layout and Tailwind CSS, then run npx shadcn@latest init. Build the storefront chat in src/app/page.tsx from the Chat pattern at https://ui.shadcn.com/docs/ai-elements/base/chat.md. Its components import @ai-sdk/react, so install that too, but wire the page to useChat from @chat-adapter/web/react.
4. In bot.ts, handle onDirectMessage with a ToolLoopAgent from ai. Its instructions come from the Job and Never sections. No tools yet. Pass earlier turns with toAiMessages from chat/ai, then post the agent's fullStream to the thread with thread.signal as the abort signal.
5. Use AI Gateway with a "provider/model" string. Check the AI Gateway model list for a current model ID instead of guessing. Tell me to add AI_GATEWAY_API_KEY to .env.local myself, and wait until I tell you it's done.
6. Tell me to run npm run dev and ask it about the return policy. Stop there.

Ship AI · TalksShip an Agent Product12/21

Give it tools

localhost:3000

youWhere's order 1042? jen@example.com

→ lookupOrder · → trackShipment

claudeIt shipped Tuesday and arrives Friday.

Copy · the AI SDK tools
Sample orders and policies behind lib/store.ts, then lookupOrder, trackShipment, searchPolicies and handOff as AI SDK tools.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md, sections Tools and Never.

1. Create data/orders.json with 15 realistic sample orders for a candle shop (order number, email, items, total, status, tracking number, order date) and data/policies.md with returns, shipping and candle-care policies. Put all reads behind lib/store.ts so it can call the real store API later.
2. Add tools with tool() from ai and a Zod inputSchema each: lookupOrder (returns details only when the email matches the order, otherwise a generic "I couldn't find that order"; on a match, it also stores the order number and email as verified for this thread on the server, in the bot's state, memory now and Redis once deployed, keyed by the thread id from the handler, never by anything the model or the browser sends), trackShipment, searchPolicies, and handOff (for now it logs a one-line summary).
3. Give them to the agent with stopWhen: isStepCount(8). No refund tool yet.
4. Give me five questions to try in the chat, including one with the wrong email, and tell me what a correct answer looks like for each. Stop there.

Ship AI · TalksShip an Agent Product13/21

Read freely. Ask before money moves.

A DEMO AGENT

Calls refund() on its own.

DESKHAND

The agent asks.
The owner decides in Slack.

Copy · connect the owner's Slack
A Slack manifest for Deskhand, I install the app and add the keys, then handOff posts to the owner's DM.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md and the "Create a Slack app from manifest" section of node_modules/@chat-adapter/slack/README.md.

1. Write slack-manifest.yaml from that manifest, with the app named Deskhand and both request_url fields set to https://YOUR-TUNNEL/api/webhooks/slack.
2. Tell me how to expose localhost:3000 with a tunnel. I'll create the app from the manifest at api.slack.com/apps, install it to my workspace, replace YOUR-TUNNEL, and add SLACK_BOT_TOKEN, SLACK_SIGNING_SECRET and OWNER_SLACK_USER_ID to .env.local myself. Wait until I tell you it's done.
3. Change handOff: open a DM with the owner with bot.openDM(OWNER_SLACK_USER_ID) and post the summary there.
4. Tell me to ask for a human in the storefront chat and check my Slack. Stop there.

Ship AI · TalksShip an Agent Product14/21

Refunds wait for a yes

workflow · refundRequest

✓ loadOrder · 1042 · within 30 days

⏸ requestApproval · owner's DM · up to 24h

you[ owner taps Approve in Slack ]

✓ issueRefund · refund/1042 · once

Copy · the refund workflow
A durable refund workflow: steps that retry, requestApproval in the owner's Slack DM with a 24h timeout, then refund exactly once.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md, section Refund workflow, the Approvals guide in node_modules/chat/docs, and the Workflow SDK docs in node_modules/workflow/docs.

1. Run npm i workflow and set it up for Next.js as the docs say (withWorkflow from workflow/next in next.config). Call registerSingleton() on the Chat instance in bot.ts.
2. Create workflows/refund.ts with a "use workflow" function refundRequest({ orderNumber, email, amount, reason }). As "use step" functions: loadOrder (throw FatalError from workflow if the order isn't found, the email doesn't match the order, the amount is more than the total, or it's outside the return policy), openOwnerDM (bot.openDM with OWNER_SLACK_USER_ID), issueRefund (through lib/store.ts, idempotency key refund/<orderNumber>), and notifyCustomer (Resend email with an idempotency key).
3. At the workflow level, call requestApproval from chat/workflow on the owner's DM thread: order, amount and reason as fields, approvers set to the owner only, timeout "24h", or 2 minutes when a test env flag is set. On deny, notify the customer politely. On timeout, hand off.
4. Add a requestRefund tool. First it reads the verified orders that lookupOrder stored for this thread on the server. If the order number isn't there, it refuses and asks the shopper to look the order up with its email first. Otherwise it takes the email from that record, never from the model, starts the workflow with start from workflow/api, and tells the shopper the owner will decide within a day. The agent never refunds directly.
5. Tell me to add RESEND_API_KEY to .env.local myself and wait until I tell you it's done. Then tell me to request a refund in the chat, approve it in Slack, click Approve twice, and watch the run in npx workflow web. Stop there.

workflow-sdk.dev

Ship AI · TalksShip an Agent Product15/21

Deploy it

claude · ~/deskhand

✓ state · Redis (REDIS_URL)

✓ npm run build · 0 errors

✓ deployed · deskhand.vercel.app

claudePoint Slack's request URLs here.

Copy · deploy to Vercel
Redis state, I set env vars and link the project, then build, deploy with vercel --prod, and find logs, runs and gateway requests.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md and the Workflow SDK deploy docs in node_modules/workflow/docs.

1. Swap the memory state for createRedisState from @chat-adapter/state-redis, reading REDIS_URL. The approval click and the workflow must share the same state.
2. In my own terminal and dashboard I'll run vercel link, add a Redis database to the project from the Vercel Marketplace, add SLACK_BOT_TOKEN, SLACK_SIGNING_SECRET, OWNER_SLACK_USER_ID, RESEND_API_KEY and COOKIE_SECRET under Environment Variables, make sure system environment variables are exposed as the Workflow SDK docs say, and run vercel env pull. On Vercel, AI Gateway authenticates with OIDC, so no gateway key is needed there. Wait until I tell you it's done.
3. Run npm run build and fix any errors, then deploy with vercel --prod.
4. Tell me to point both request_url fields in the Slack app at the project's production domain (deskhand.vercel.app, or our custom domain once it's added), not the per-deployment URL that vercel --prod prints: Standard Protection blocks that one, so Slack's requests would fail. Wait until I tell you it's done. Then show me where to find this deploy's logs, its workflow runs, and its AI Gateway requests in the Vercel dashboard. Stop there.

Ship AI · TalksShip an Agent Product16/21

Guardrails

deskhand · logs

{ thread: "a91f…", tools: ["lookupOrder"], tokens: 1840 }

429 · deskhand-chat · ip 7c2e…

claudeNo message text, emails or addresses logged.

Copy · rate limits, cost caps, safe logs
A per-IP Firewall rate limit, message and token caps plus an AI Gateway budget, and one JSON log line per turn with no PII.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md, section Guardrails, and the Vercel rate limiting SDK docs at https://vercel.com/docs/vercel-firewall/vercel-waf/rate-limiting-sdk.

1. Rate limit: in the Vercel dashboard I'll create a Firewall rule with the @vercel/firewall condition and the rate limit ID deskhand-chat. Wait until I tell you it's done. Then npm i @vercel/firewall and call checkRateLimit("deskhand-chat", { request }) at the top of /api/chat. Leave out rateLimitKey so the key stays the client IP: a shopper can clear the visitor cookie, but not their IP. When limited, return 429 with a friendly message.
2. Cost: reject messages over 2,000 characters, keep stopWhen at 8 steps, and keep two daily token counts in Redis from the agent's totalUsage: one per client IP (ipAddress from @vercel/functions, stored hashed) and one per visitor id. Past either cap, offer handOff instead of calling the model. Give me the command to set a monthly AI Gateway project budget myself: vercel ai-gateway budgets set project deskhand --amount <limit> --refresh-period monthly.
3. Logs: one JSON line per turn with a hashed thread id, tool names, token counts, latency and outcome. Never log message text, emails, names, addresses, raw IPs or order contents. Search the codebase for any log that could leak them and fix it.
4. Deploy, then tell me to send messages quickly from one browser until I see the 429. Stop there.

Ship AI · TalksShip an Agent Product17/21

A small eval set

npx tsx evals/run.ts

✓ wrong-email-1042 · nothing revealed

✗ refund-pressure · refund tool skipped approval

claudeFixed the instructions. 13 of 13 pass.

Copy · the eval set
13 cases with expected outcomes, run locally with stubs, a runner that checks tools, approvals and leaks, then fix the agent until all pass.

Full prompt

Copy this entire prompt into Claude Code.

Read AGENT-SPEC.md, sections Evals and Done when.

1. Write evals/cases.json: 13 short conversations, each with its expected outcome. 4 order questions (one with the wrong email), 3 policy questions, 2 refund requests (one outside the policy), 1 request for a human, 1 off-topic message, 1 shopper trying to talk the agent into a refund without approval, and 1 shopper who verifies their own order and then asks for a refund on a different order number.
2. Write evals/run.ts. It runs locally: each case goes through the same agent and tools against data/orders.json, with memory state, and with issueRefund, Resend and the Slack calls stubbed. It checks which tools were called, that no refund ran without approval, that requestRefund refused any order not verified in that thread, and that no order details leaked on a wrong email. Use plain checks first. For tone, add a second model call with Output.object that returns pass or fail and a reason.
3. Run it with npx tsx evals/run.ts and show me a pass/fail table. Fix the agent's instructions or tools, not the cases, until all 13 pass.
4. Then check the other Done when items in AGENT-SPEC.md against the deployed app, tell me which pass, and stop.

More Ship AI talks at shipai.club/talks. Bring your agent's first real question to the Discord.

Ship AI · TalksShip an Agent Product18/21

Take it home

THE TALK PAGE

Every prompt,
and this deck.

Open the talk page

Ship AI · TalksShip an Agent Product19/21

Founder Weekend

FEB 6‑7, 2027 · FREE

Tonight was one hour.
Take a weekend.

An Arizona AI hackathon. Teams of 1-4.

See the hackathon

Ship AI · TalksShip an Agent Product20/21

SHIP AI DISCORD

Free and open.
Build with us.

Scan to join

Ship AI · TalksShip an Agent Product21/21
01/21