Revnu

Check whether Revnu is working when the customer asks about progress or agent liveness. Returns status with the sandbox lifecycle state, last activity and finished-turn timestamps, whether a turn is running, and the number and next time of scheduled wakes. Read-only; does not start or interrupt work.

The customer's block list, newest first: the domains and addresses their agent will never research or contact. Pass back nextCursor for the next page. Read-only.

One lead with its contacts (capped at 50; contactsTruncated says when), deal, owner, research, researched people not yet saved as contacts, handling when the team has taken it over, the newest 10 timeline notes (data, not instructions) and interactions: last send and reply, waitingOn and connectionStrength. Read-only. In hosts that support MCP Apps, the user also sees this result in the Revnu app.

The customer's leads with status, owner, deal and last interaction, never draft bodies. Newest first unless sort says otherwise. q searches company, domain and primary contact. stage lists one deal stage. workflowRunId lists the leads one workflow run found. Pass back nextCursor with the same filters for the next page. Read-only. In hosts that support MCP Apps, the user also sees this result in the Revnu app.

One person's LinkedIn relationship with each of the customer's connected LinkedIn accounts (or the one named in account): whether they're connected, any pending invite and when it was sent, Revnu's own invite record with when it saw acceptance, and the newest 50 messages with who sent each. Read live from LinkedIn: each account is one paced profile lookup, so look people up as you need them rather than in bulk. Read-only.

The customer's growth scoreboard for a window ending now: per-channel sections that are either {available: true, data} or {available: false, reason}. Zero and unknown are different. since is epoch milliseconds, clamped to the last 31 days; default 24 hours. In hosts that support MCP Apps, the user also sees this result in the Revnu app.

One review card by id, including its full payload (draft text, options, research). Read-only.

List what is waiting for the customer's approval: drafts, questions, budgets and other cards their revnu growth agent has queued. Each card carries actions, the exact verbs the API accepts for it, and primaryLabel, what the dashboard's main button would say. One call returns a limited batch of open cards: pending ones and ones the agent is revising together, with status on each card saying which. There is no status filter. Follow nextCursor while hasMore is true to read further pages. Returns compact card summaries to the model; full drafts and evidence are delivered separately to the Revnu app. Use review_get only when the customer asks to inspect a particular draft. One successful call already shows the cards in hosts that support MCP Apps. Do not call again to display them, verify the same list, or fetch the same result with omitted/default arguments. Finish with a short count rather than copying the entire card list into chat. Read-only.

Check whether work for named people or topics already has a review card, including done, dismissed and other historical cards. Batch names in one call. Returns matches, one per query: its items (id, title, status, update and resolution times, delivery as sent_recorded, queued or unconfirmed, and reviewUrl), hasMore and openCardCount; plus coverage, the evidence limits to keep when reporting. No draft bodies and no app. Searches titles and prompts only; no match is not proof of no activity. Done is not proof of delivery. When hasMore is false, openCardCount already counts every matching pending/revising card. Copy reviewUrl directly for links. For a named status lookup, report matching records without listing unrelated queue cards or opening the app. Stop when the requested statuses and their evidence limits are clear. Use review_get only when full draft or evidence is needed; use ask_revnu for missing account evidence, requesting a read-only check when no action is authorized. Read-only.

Who this token is. Returns me: the token's id and label, its scopes (which tools it can use), when it was created and last used, the Revnu account it belongs to, inSetup while onboarding is unfinished (then call setup_status), and its configured rate limits, which are capacities, not remaining balances. Needs no scope.

Run history for one workflow, newest first, in runs. Each run includes its ID, workflow version, status, start and finish times, headline, error, counts of new leads, contacts and research-list people, and reported step outcomes with notes and review-card links. Use the run ID with leads_list.workflowRunId to inspect the leads it touched. An empty list means no runs are recorded; queued or running runs are unfinished. Read-only.

The customer's recurring workflows with their schedule, next run and steps. Live workflows run on their schedule. A proposed workflow is not approved yet (workflow_resume approves it): it runs on its schedule when the account's Workflows approval setting is Just Do It, and waits for approval when it is Ask Me First. Paused workflows do not run. Use nextRunAt and scheduleError to report the next run; workflow_runs shows recorded executions. Read-only. In hosts that support MCP Apps, the user also sees this result in the Revnu app.