Welcome back
Your pre-send activity at a glance.
AI usage
AI checks on your plan this month
Quick test
Try the checker right here — no inbox needed
Rules
Your pre-send checks, per platform. Risky messages are stopped at the gate before they leave.
Checks run in two tiers: instant local checks fire while you type, while AI and connected-tool checks run once at send time — so deep checks never slow the keyboard down.
Built-in checks
Instant local checks — zero AI cost, zero latency
Actions
Small workflows that run when you hit Send. Each step does one thing: call your own API, ask a connected tool, ask the AI, match a pattern. A Decide step then blocks, warns or notes based on what the steps found.
Actions run once, at send time, never while you type. A step that fails is skipped, so an outage on your side never blocks a send.
Approvals
Messages held until someone signs them off. Give a rule approvers and Presend asks them before the message can leave, in any app it runs in.
Waiting on you
Messages other people want to send that name you as an approver
Your requests
Messages you asked to have approved, newest first
Connections
The external tools Presend can consult at send time. Connect an MCP server or a plain REST API and your rules can check a message against the systems you already run on.
Voice
Teach Presend how you actually write, so it can score and rewrite in your tone — not ChatGPT's.
Your voice profile
—
Add more real sent messages to sharpen the match.
Saved samples
Billing
Manage your plan and usage.
Usage this period
—
Settings
Connection, AI configuration and account.
Connection
Where this dashboard talks to
API keys
For the REST API and the MCP server
A key does not expire, unlike your browser session, so an agent or a script keeps working. Treat it like a password: it can check messages as you.
AI configuration
How AI checks are powered
Stored server-side and only used for your requests.