Where the product loop is, and what to do next

product.statusReadsOnly looks things up; never changes your workspace.

What it does

The product-level next step, the twin of pdlc.verify's nextStep but for the whole product rather than one feature. Fixed rules over stored records, no model involved. Returns nextStep {action, reason, prompt, tool, who}: "assess" = no strategy assessment yet, or the strategy documents changed since the last one (run the product-strategy prompt, then product.assess); "review_assessment" = the latest assessment is waiting for a person on the Dashboard (nothing for an agent to do unless changes were requested); "set_goals" = no active product KPI (propose goals with productkpi.create); "fill_backlog" = no feature yet (feature.implement); "measure" = an active KPI has no observation in 30 days (metric.observe); "none" = nothing waiting. Also returns the latest assessment (review status, right-to-win, stale flag), goals, features and stories counts. Pass manifest (the files object from npx @testmaze/mcp docs manifest) to learn whether the reviewed assessment is stale against the documents on disk; without it staleness is reported as null. kind is the workspace kind: "product" runs this loop; "service" (a client engagement) tracks delivery metrics only, so its next step is fill_backlog or none and product.assess / productkpi.* are refused. Read-only. Call it after project.whoami, and again after product.assess or a review.

Ask your agent

You don’t call product.status yourself. Say something like this to Claude Code, Cursor or another MCP-connected agent:

  • “What should we do next for this product as a whole?”
  • “Is the strategy assessment still current, or did the docs change?”
  • “Do we have product goals yet?”

Inputs

NameTypeDescription
manifest
optional
objectOptional: the files object printed by npx @testmaze/mcp docs manifest ({"<path>": {md5, bytes}}), to compare the documents on disk with the ones the latest assessment cited.

The MCP call

What the agent’s MCP client sends (placeholders in angle brackets):

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "product.status",
    "arguments": {
      "manifest": {}
    }
  }
}

Connect your agent

npx -y @testmaze/mcp init tmt_xxx
claude mcp add tm --scope project -- npx -y @testmaze/mcp

Create the token in your Test Maze workspace under Settings → MCP. Setup for Cursor, Cline, Gemini CLI and Codex CLI is shown there too.