How ONE table is doing, its Overview. All time: the flow of its people and of its accounts over seven milestones (in_table, reached, engaged, replied, positive, meeting, won), each with how many got at least that far, who sits at it by current stage (here) and who left from it by exit stage (left); the money on its accounts by deal size (open pipeline, won, lost, the accounts without a size counted); its data sources with how far their people got. A person counts at every milestone up to the furthest the ledger or the stages say they reached (a stage stands for its category's milestone; a nurture, negative or disqualified stage is an exit), so the flow narrows by construction. For the window asked for (days ending today, or a from..to range) and the one before it, day by day in the workspace's zone: people added, reached, replied, turned positive, won, and the deal size won; sum each half and compare. Use for "how is this table doing", "where do people drop out", "how much is in the pipeline", "which source works". What happened to whom: search_activities with filter.playtable_sid.

Contract:

• MCP tool get_playtable_performance, registry package mcp.playtables/playtables, mount playtables.tables.
• Operation action, response envelope action.
• Flags: read only.