Read jobs, customers and other records from Fieldproxy

Read by exact match#

An administrator creates a key in the web app under Settings, Security, API keys: give it a name, an expiry, only the operations this guide needs, and optionally only the tables it touches. Keep it in an environment variable, FIELDPROXY_API_KEY. Your API base is your workspace's regional host, for example https://api-india-1.fieldproxy.ai (India) or https://api-us-east-1.fieldproxy.ai (US East). This guide needs the Read operation.

condition is a non-empty object of column-value pairs. Every pair must match, by equality. It cannot be empty, so a call can never return or touch a whole table by accident.

curl -X POST https://api-india-1.fieldproxy.ai/api/read \
  -H "x-api-key: $FIELDPROXY_API_KEY" \
  -H "content-type: application/json" \
  -d '{"tableName":"jobs","condition":{"status":"assigned","technician_id":42}}'

// Node.js 18+ (built-in fetch)
const res = await fetch("https://api-india-1.fieldproxy.ai/api/read", {
  method: "POST",
  headers: { "x-api-key": process.env.FIELDPROXY_API_KEY, "content-type": "application/json" },
  body: JSON.stringify({
    "tableName": "jobs",
    "condition": {
      "status": "assigned",
      "technician_id": 42
    }
  }),
})
const out = await res.json()
if (!res.ok) throw new Error(out.message)
console.log(out.rowCount, out.data)

# Python 3 (pip install requests)
import os, requests

res = requests.post(
    "https://api-india-1.fieldproxy.ai/api/read",
    headers={"x-api-key": os.environ["FIELDPROXY_API_KEY"]},
    json={
      "tableName": "jobs",
      "condition": {
        "status": "assigned",
        "technician_id": 42
      }
    },
    timeout=30,
)
out = res.json()
res.raise_for_status()
print(out["rowCount"], out["data"])

{
  "data": [ { "id": 1042, "title": "Boiler service", "status": "assigned", "technician_id": 42, "...": "every column" } ],
  "rowCount": 1,
  "table": "jobs",
  "tenant": "<workspace-id>"
}

Exact matches here, anything else with SQL#

• /api/read matches values exactly. For ranges, IN lists, patterns, joins or totals, use read-only SQL (below).
• Every matching row comes back with all its columns in one response (a test read of 5,001 rows did), so make the condition specific. For a sorted or trimmed list, use SQL.
• An unknown column returns 500 with Read failed and the database error.

Ranges, joins and totals: read-only SQL#

Connect an MCP client (Claude, ChatGPT, Cursor, Claude Code or your own agent) and use run_sql: one read-only SELECT on the workspace's own Postgres database, up to 500 rows, with a 5-second limit. export_csv saves up to 10,000 rows as a CSV link.

describe_schema({ "tables": ["jobs", "customers"] })

run_sql({ "sql": "SELECT j.id, j.title, c.company_name, j.status
  FROM jobs j JOIN customers c ON c.id = j.customer_id
  WHERE j.status = 'completed' LIMIT 50" })

Table and column names differ per workspace, because each one is set up around its own business. The names below (jobs, customers, invoices) are typical, not guaranteed. To see the real ones, connect any MCP client and call describe_schema, or ask your Fieldproxy administrator.

Questions#

Checked against the Fieldproxy platform reference on October 18, 2026. Plain text for AI tools: /developers/guides/read-jobs.md