Skip to Content
Getting StartedQuickstart

Quickstart

This walks the full loop a typical integration follows: authenticate → store a resume → stream an AI edit. Replace localhost ports with your production base URL as needed.

1. Mint a Personal Access Token

export CLERK_JWT="<clerk-session-jwt>"
 
curl -s -X POST http://localhost:3111/api/users/me/personal-access-tokens \
  -H "Authorization: Bearer $CLERK_JWT" \
  -H "Content-Type: application/json" \
  -d '{ "name": "quickstart" }' | jq -r '.data.token'

Save the mr_pat_… value:

export PAT="mr_pat_xxxxxxxxxxxxxxxxxxxxxxxx"

2. Create a resume

content is a serialized JSON string of your resume document — the backend stores it opaquely.

curl -s -X POST http://localhost:3111/api/resumes \
  -H "Authorization: Bearer $PAT" \
  -H "Content-Type: application/json" \
  -d '{
    "title": "Frontend Resume",
    "content": "{\"basics\":{\"name\":\"Ada Lovelace\"},\"work\":[]}"
  }' | jq '.data.id'

3. List your resumes

curl -s http://localhost:3111/api/resumes/mine \
  -H "Authorization: Bearer $PAT" | jq '.data'

4. Stream an AI chat

The chat endpoint streams Server-Sent Events. Each data: line is a JSON object with a type. Watch for message_chunk (incremental text) and resume_update (a full revised resume).

curl -N -X POST http://localhost:3112/api/chat \
  -H "Content-Type: application/json" \
  -d '{
    "mode": "optimize",
    "messages": [
      { "role": "user", "content": "Make my summary punchier and quantify impact." }
    ],
    "currentResume": { "basics": { "name": "Ada Lovelace" } }
  }'

You’ll see a stream like:

data: {"type":"run_started","runId":"clx..."}
data: {"type":"message_chunk","content":"Here","runId":"clx..."}
data: {"type":"message_chunk","content":"'s a tighter","runId":"clx..."}
data: {"type":"resume_update","data":{...revised resume...},"runId":"clx..."}
data: {"type":"run_completed","runId":"clx..."}
data: {"type":"done","runId":"clx..."}

In TypeScript

A minimal client that wraps the envelope and consumes the SSE stream:

const PLATFORM = 'http://localhost:3111'
const PAT = process.env.PAT!
 
async function api<T>(path: string, init: RequestInit = {}): Promise<T> {
  const res = await fetch(`${PLATFORM}/api${path}`, {
    ...init,
    headers: {
      Authorization: `Bearer ${PAT}`,
      'Content-Type': 'application/json',
      ...init.headers,
    },
  })
  const body = await res.json()
  if (!res.ok || body.code >= 400) {
    throw new Error(body.message ?? `HTTP ${res.status}`)
  }
  return body.data as T // unwrap the envelope
}
 
const resumes = await api<Array<{ id: string; title: string }>>('/resumes/mine')

SSE frames are separated by a blank line (\n\n). Buffer across reads and only parse complete frames — a single network chunk may split a frame in half. See SSE events for the full event catalog.

Next steps

Last updated on
AuthenticationOverview