Prysm
// Module — Connections

Your data, already where you work.

Plug Postgres, MySQL, REST APIs, Google Sheets, Excel, and files straight into PrysmOS. Hosts are validated, credentials stay encrypted, every connection is tested before it saves — then tables land as real spreadsheet tabs that sync on your schedule.

  • Postgres, MySQL, REST, Sheets, Excel & files
  • Credentials encrypted — never echoed back
  • Scheduled & live sync, pull or push
  • Org file storage your scripts can read

prod-postgres

postgresql://db.acme.co:5432

tested · connected
public.customersid · email · lifetime_value
public.ordersid · status · amount · ts
public.eventsid · type · payload
auto-pull · nightly 02:00write-back · oncreds encrypted

browse every table & column before a single row lands

// How it works

Connected in minutes, fresh forever

No pipeline code, no cron jobs to babysit — a connection is a living thing that syncs on its own.

01

Pick a connector

PostgreSQL, MySQL, any REST API, Google Sheets, Microsoft Excel, or drop a file. Six connectors cover where team data actually lives.

02

Test it before it saves

PrysmOS validates the host — unsafe and internal targets are blocked — then test-fires the connection. It only saves once it actually connects.

03

Browse before you import

Schema introspection shows every table, column, and endpoint up front. Choose exactly what lands — no all-or-nothing pulls.

04

It keeps itself fresh

Scheduled or live sync, pull or push on demand. External tables materialize as real spreadsheet tabs — and write-back sends clean rows home.

// What's inside

Six connectors, zero plumbing

The boring parts — credentials, validation, scheduling, schema browsing — are already done. You just pick what to pull.

PostgreSQL & MySQL

Point at the database, browse every schema, and pull the tables you actually need — nothing else.

REST APIs

Any JSON endpoint becomes a queryable source. GETs pull on a schedule; POST and PATCH targets enable write-back.

Google Sheets

OAuth in, pick a workbook, and its tabs sync into PrysmOS — the source of truth stays where the team keeps it.

Microsoft Excel

OneDrive and SharePoint workbooks connect over OAuth too. The budget file updates itself — no more email attachments.

Files as live sources

CSV, XLSX, and JSON drop straight in and become queryable like any table. Re-dropping a file refreshes its data.

Credentials stay sealed

Secrets are encrypted at rest and never echoed back — not to logs, not to the UI, not to anyone. OAuth means we never see a password at all.

Host validation built in

Unsafe targets are refused before a connection can save — internal ranges, link-local addresses, and known-bad hosts never get reached.

Test before you save

Every connection runs a live test against the source first. If it can't connect, it doesn't save — no broken sources limping along.

Schema introspection

Browse tables, columns, and endpoints before importing. Preview the shape of the data, then choose exactly what lands.

External tables, real tabs

Connected tables materialize as spreadsheet tabs — live recalc, pivots, charts, and contracts all work on day one.

Scheduled & live sync

Sync on a cadence — every 30 minutes, hourly, nightly — or keep sources live. Pull fresh data or push edits whenever you want.

Org file storage

A shared file space for the whole org. Scripts read and write it, Juno can pull files into its answers, and anything in it is a potential data source.

// Try it

Pick a source, browse its schema

Click a connector to connect it, then expand the tables — this is the same introspection you'd see before a real import.

app.prysmhq.co/connections

Add a connection

Credentials are encrypted at rest. OAuth sources never share a password.

Pick a connector on the left — you'll browse its schema before a single row imports.

// In the wild

Connections that delete busywork

auto-pull

The nightly CRM pull

The CRM's REST API syncs every night at 02:00. The morning standup sheet is already fresh — nobody exports anything, nobody asks which version is current.

write-back

Clean rows, sent home

The ops team dedupes and fixes 400 Postgres rows in the grid — real formulas, flash-fill, the works. Write-back pushes the clean set upstream. The database stays the source of truth.

no attachments

Kill the emailed Excel file

Finance's budget workbook lives on OneDrive. Connect it once over OAuth and it syncs nightly — 'budget_v7_FINAL_final.xlsx' never hits an inbox again.

// Questions

Fine print, answered

Six: PostgreSQL, MySQL, REST APIs, Google Sheets, Microsoft Excel, and file uploads (CSV, XLSX, JSON). Together they cover databases, SaaS endpoints, cloud workbooks, and everything that still arrives as an attachment — including files already in your org's storage.

// One workspace

Stop moving data by hand.

Connect a source once — it syncs on a schedule, lands as a real spreadsheet, and writes clean rows back. Free to start — no credit card required.