MCP Tool Playground

Invoke each MCP tool against your workspace's seed data. recommend_play runs the real matcher; the paid tools are gated. Gating here is a client-side prototype — real enforcement is a server-side check on the tool call.

get_standard()Free

Returns the five-object Standard schema (Company, Account, Person, Signal, Engagement), versioned, so the agent can interpret whatever a vendor tool returns.

get_vendor_mapping(vendor)Free

The field-mapping table for a connected vendor (Apollo first) so the agent can reconcile that vendor's data against the Standard.

get_workspace_context()Free

Read-only, capped access to the workspace's approved context — profile, ICPs, personas, and active plays.

list_templates() / get_template(id) / push_template(id, target_tool)Free

Browse and format the prompt/template library for the destination AI tool (Claude, ChatGPT, Gemini, Cursor).

recommend_play(signal)Free/Paid

Matches a signal to an approved play. Free tier resolves against the public play library; paid tier also resolves against the workspace's own approved plays.

Paid feature

generate_campaign_brief is available on the Workspace plan and above.

Upgrade →

Paid feature

translate_object is available on the Workspace plan and above.

Upgrade →

How gating works here

Toggle the tier above. On the Free tier, recommend_play returns only public plays, and generate_campaign_brief / translate_object are gated. This is a client-side prototype of the gate — in production the MCP server rejects a free-tier token calling a paid tool.