Skip to content
SynthBoard

Concepts

Actions with receipts

Advice that stops at advice leaves the follow-through to you. With integrations connected, SynthBoard can do the follow-up — write up the page, open the ticket, draft the reply — and every write waits for your confirm. Every tool call leaves a receipt.

62
Providers
Across 22 categories — Gmail, Slack, Linear, Notion, HubSpot and more.
You
Confirm every write
A preview card first. Nothing lands until you confirm it.
Pro+
Writes
Reading your tools is never gated, on any plan.

Writes from your Chief of Staff#

On Home, your Chief of Staff can write into a connected tool for you. She never writes on her own: every write is a card you confirm.

  1. She prepares one of five kinds of write: a new Notion page, a new Google Doc, a Slack message, a Linear issue, a Gmail draft (you send it yourself).
  2. It arrives in the thread as a preview card — where it goes and what it says — with a confirm button. Nothing is written until you press it; dismiss the card and nothing is written at all.
  3. When you confirm, the write is re-checked from the card itself, never from what the page sends: your plan, the connection, then the same audited gateway every tool call goes through.
  4. If the tool is not connected yet, the card lets you connect it in place and then confirm.
  5. She drafts from work already done and paid for. If a write would need new judgment, she offers a room instead of a card.
Errands
An errand — "go over my Notion and write up what changed" — ends the same way. She reads what she needs, says what she read, prices only the part that needs an expert, and prepares the page from that expert’s delivered work. The page opens with a “Needs your review” note and lands as a new page only when you confirm.
Plans
Writes are on Pro and above. On Free she says so and gives you the words to paste yourself — no card. Reading your connected tools is never gated, on any plan.

The one write that does not wait for a card is a share rule you set up yourself in settings — post each finished board’s headline and link to a Slack channel or a Notion log. It runs until you turn it off, and never for a room that is off the record.

Actions proposed inside a room, the API and MCP#

Actions an expert proposes inside a session, or that you run over the API or MCP, go through the older governance pipeline, which is still in place. Three gates stand between an expert's idea and anything actually happening.

  1. Propose. An expert proposes the action inside its response — grounded in the context it just argued from (the Gmail thread it read, the Linear ticket it cited).
  2. Your approval. You see a clean preview of exactly what will run — recipient, content, target record — and nothing executes until you approve it.
  3. Execute, with a receipt. The action runs once, through the same audited gateway as every tool call, and the result is logged with a link to what was created.

The receipt#

Every executed action leaves a record you can audit.

Each action that runs is logged with its full provenance: which expert proposed it, when you approved it, and what actually executed. The session record keeps these receipts alongside the memo — so the decision and its follow-through live in one place. Her writes from Home leave the same audit record.

Why receipts matter
When an AI can touch your email, your CRM, and your issue tracker, “trust us” is not a governance model. A receipt for every action — plus your confirm before anything lands — is.

Your controls#

Execution is opt-in and scopeable, per session.

  • Disable per session — run any session with tool execution off entirely.
  • Restrict by tool — allow only specific tool slugs (for example: Gmail drafts only, Linear reads only). Over the API and MCP the same control ships as integrations_enabled and allowed_tool_slugs on the session config.
  • Reads vs. writes — reading context (a thread, a ticket, a deal) and proposing an action are separate steps; connecting an integration never grants silent write access.

Where to go next#

Actions sit on top of the integration and memory layers.

  • Integrations — the 62 providers your board can read from and act on.
  • Security posture — how tool-execution governance fits the wider trust model.
  • Your World, with sources — what she keeps on file, which makes proposed actions contextual instead of generic.