Capability matrix
Campaign templates/briefs/objections, signals, mailbox health, reply tagging, sequence pause/resume and outbound statistics are also available through
POST /workspace/actions and MCP workspace_action. See workspace actions for action fields and confirmation behavior.
What this means for AI agents
Discovery and intelligence actions reuse the app’s real handlers, identity checks, credits and limits. A typical agent workflow is:- Find new prospects with
search_prospects/POST /prospects/search; these are not saved automatically. - Check an address with
verify_email/POST /email/verify. Onlyverified: trueconfirms a mailbox; catch-all, role and unknown results remain unproven. - Save approved prospects with
create_lead/POST /leads, then enrich, research, score and draft. - Ask the user before enrolling a lead or sending a reply.
search_leads still searches saved CRM records; search_prospects searches the external B2B network. Do not invent results or treat a failed request as an empty successful search. The table above is the current contract, not a claim that every app settings/admin screen is available remotely.
Credits and plan limits
- API and MCP access require a paid plan (Starter and above). See errors and limits for request allowances.
- Enrichment and email generation use the same account entitlements as the app. An API key never bypasses plan limits.
- Request allowances (per minute / per day) and feature credits are separate balances.
- Paid prospect searches cost no reveal credits but count against the same daily search meter; results are capped by plan (Starter 25, Growth 50, Pro 100, Agency 200).
- Person lookup and saved-lead enrichment retain the app’s one-credit reserve/refund behavior.
- Standalone verification charges no reveal credit, with 10 requests/minute and rolling daily checks capped at your plan’s daily search limit. API, MCP and agent chat share this verification allowance.