ViablyViably
Cursor

Debug the webhook with the payload in hand.

Add Viably to Cursor and the agent can read what your relay endpoints received, file the bug it finds as an issue in Slack, and check the pipeline while you stay in the editor. It works as you, and it asks before every change.

Add to SlackRead the setup guide
Cursor

why are the Stripe webhooks failing since this morning?

viably.query_relay_logs (status: "failed", type: "stripe")12 failed deliveries since 09:40, all charge.succeeded

All twelve fail the signature check, and they start with the commit that moved the body parser above the webhook route. Stripe signs the raw body, so the route needs it before JSON parsing. Want me to file this in #engineering?

Cursor wants to run create_issueDenyAllow
Connect itAdd the server to mcp.json, globally or for one project, and approve the sign-in Cursor opens
What you needA Viably workspace you belong to. The agent gets your role, so it sees what you see
Up and running in2 minutes
On your planEvery plan, with 500 tool calls a month on Free and 15,000 on Viably+
Where it landsYour inbox, pipeline and board, as if you had made the changeThe Slack card of any task or issue it filesThe audit trail on API and agentsTwo-way
What it can reach

Your workspace, as tools an agent can call.

The server lists these to any app that connects. Reads are ticked when you sign in, and each area of changes needs its own tick.

Inbox and contacts

A list shows names only to a member who may also download contacts, since an agent can page through the whole book.
list_inboxfind_contactsget_contactclaim_contactset_contact_statustag_contactadd_contact_notedismiss_inbox_itemsnooze_inbox_item

Deals

Stages come from your own pipelines, and list_deals returns the keys move_deal expects.
list_dealsopen_dealmove_deal

Tasks, issues and the calendar

A task or issue an agent creates posts its card to Slack, the same as one filed from the dashboard.
list_taskscreate_taskupdate_tasklist_issuesget_issuecreate_issueresolve_issuelist_eventscreate_eventset_reminder

Logs, search and signals

Replaying a relay event is for workspace admins.
query_relay_logsreplay_relay_eventsearch_console_summarysocial_summaryuptime_statusgithub_activitymedal_leaderboardsearchwhoami
In the editor

The event log, one question away.

query_relay_logs searches every event your endpoints received, with the payload each one arrived in. Ask why a webhook failed and the agent reads the failures next to the code that handles them. If you are an admin and grant relay changes, it can replay those events once the fix ships.

ViablyViably
The Viably webhook relay, showing endpoint health, event volume over ninety days, and the searchable delivery log
How the relay works
Per project

One config file, a sign-in per person.

Put the server in the project's .cursor/mcp.json and commit it. Everyone who opens the project gets the same entry and signs in as themselves, so each person's agent carries their own role and the scopes they ticked.

  • Global or per project

    Add the server to ~/.cursor/mcp.json for every project, or to a project's .cursor/mcp.json to share it with the team through the repo.

  • Only the URL in the repo

    The file holds only the server URL. Access comes from each person's sign-in, which Cursor renews on its own.

  • Leaving ends it

    Viably checks the member on every call. When someone leaves the workspace, their sign-ins stop working right away.

viably.app/oauth/authorize

Connect Cursor to Viably

Cursor will act in your workspace as you, limited to what your role allows. Untick anything it doesn’t need.

What it can read in Athlead

Your inbox

Contacts

Deals

Tasks

Issues

What it can change

Claim contacts, set their status, tag them and add notes

Open deals and move them between stages

Create tasks, which post to Slack, and change their status, assignee and due date

Create issues and resolve them, noting it in their Slack thread

AllowCancel
Filing

The bug you just found, filed where the team looks.

Ask the agent to file it as an issue in #engineering, or as a follow-up task due Monday. The card posts to Slack under your name and lands on the board, and an issue reaches GitHub too when the channel has a linked repository.

ViablyViably
The Viably kanban board with issue and task cards grouped into open, active, and closed columns
See issue tracking

Every tool on every plan, counted by the call.

Free includes 500 tool calls a month and Viably+ includes 15,000, shared by the whole workspace. A top-up pack adds 3,000 more. Connecting a second app or a second teammate costs nothing extra.

14 days
Viably+ trial.
Seats on every plan
$7
per day, billed annually
3,000
Events each month
The Viably relay log, listing webhook deliveries with their payloads
The Viably board, holding tasks and issues filed from Slack, the dashboard and connected agents
The Viably contacts dashboard, the record an agent reads through find_contacts and get_contact
Every event your endpoints received, with the payload query_relay_logs hands to your agent.
FAQ

Questions about the Cursor integration

What goes in mcp.json?
One entry: {"mcpServers": {"viably": {"url": "https://api.viably.app/mcp"}}}. Put it in ~/.cursor/mcp.json for every project, or in .cursor/mcp.json inside one. Cursor opens the Viably sign-in the first time the agent needs it.
Can the agent replay failed webhooks?
Yes, for a workspace admin who ticks relay changes at sign-in. replay_relay_event sends a logged event to Slack again, so you can fix the handler and replay what it dropped.
Will issues I file from Cursor reach GitHub?
If the channel you file into has a linked repository, yes. The issue posts to Slack, sits on the Viably board, and mirrors to GitHub with the same title and body.
Can the agent see anything I can't?
No. Every call runs as you, through the same checks the dashboard runs. If your role hides customer names, the agent gets them masked too, and the scopes you untick at sign-in narrow its access further.
Can it email or message our customers?
No tool sends anything to a customer. An agent can leave a draft reply as a note on the contact, and a person sends it.
What if a customer's message tries to give the agent instructions?
Anyone who writes to your inbox or fills in your form puts text in front of the agent. Viably labels that text as data in every tool result, and no tool can reply to a customer. Keep your app asking before each change, and give an agent that runs unattended read scopes only.
What happens when someone leaves the workspace?
Their agents lose access on the next call. Viably checks the member behind every token each time, so there is nothing to revoke by hand.
Read the full guideConnect Claude and Other Agents (MCP)
Go deeper

Where this leads next

MCP server guide

Setup for each app, what every scope allows, the saved prompts, and the rate limits.

REST API reference

The same tools as HTTP endpoints, for scripts and CI jobs that call Viably without an agent.

More integrations

Connect something else

Everything lands in the same record, so the second service you connect takes less setting up than the first.

Claude

Claude Code and Claude.ai, working your inbox as you

Codex

Deals, tasks and relay logs, from the terminal

Slack

@Via coordinates tasks and automations in your Slack channels. Natural language support and emoji-triggered shortcuts.

Ready when you are

Keep the payload and the fix in one window.

Point a webhook at a relay endpoint and add one entry to mcp.json. The next failed delivery is one question away.

Add to SlackSee pricing
SignalRecordOwned work
  • Free to start
  • No credit card required
  • Powers on when you need them