Record a product assessment (strategy → Test Maze)
product.assessChanges dataCreates, updates or deletes something in your workspace.What it does
The last step of the product-strategy prompt. Stores what you concluded from the repo's strategy documents as a Pending assessment a person reviews on the Dashboard: summary, right-to-win (0-10, an assessment, never a KPI), reasoning, ten market estimates, spend analysis, payback / stage / recommendation and up to five action items. Provenance is enforced: pass manifest exactly as npx @testmaze/mcp docs manifest printed it AFTER your last edit, and every estimate, action item and provenance.sources entry must cite a file from it with the same md5 — a value with no source is rejected unless it is "<not stated>". Action items with createFeature (default true) become Pending features; goals become proposed KPIs (productkpi.create rules). Returns the assessment, the features and KPIs created, and the product's next step. Never call it to re-record the same documents: check product.status first.
Ask your agent
You don’t call product.assess yourself. Say something like this to Claude Code, Cursor or another MCP-connected agent:
- “Save the strategy assessment I just wrote into Test Maze”
- “Record the right-to-win analysis with its sources”
Inputs
| Name | Type | Description |
|---|---|---|
summaryrequired | string | Two or three sentences: what the product is and why it can win. |
rightToWinoptional | number | 0-10, one decimal. An assessment with its reasoning, never a gate. |
reasoningoptional | string | 3-5 sentences: ROI, timing and competition, threats, opportunities. |
estimatesoptional | object[] | The ten market estimates (one per key), each citing its source document. |
spendAnalysisoptional | string | Markdown: investment to date, phases, ROI, recommended short / medium / long-term investment, continue? and budget split. |
paybackoptional | string | e.g. "18 months" or "<not stated>". |
stageoptional | enum | Investment stage from the investment case. Allowed: conceptmvpgrowthmature |
recommendationoptional | enum | The investment case's recommendation. Allowed: approveconditionalreject |
actionItemsoptional | object[] | Up to five action items; each with createFeature (default true) becomes a Pending feature. |
goalsoptional | object[] | Product KPIs this strategy implies; saved as proposed for a person to activate. |
manifestrequired | object | Output of npx @testmaze/mcp docs manifest, run after your last edit. |
provenancerequired | object | Who produced this (recordedBy agent), the model, the session, and every document read ({path, md5}). |
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.assess",
"arguments": {
"summary": "<summary>",
"manifest": {
"files": {}
},
"provenance": {
"recordedBy": "agent"
}
}
}
}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.
More product & connection tools
- Check which workspace this connection uses
project.whoami - Get the product details for this workspace
product.get - Where the product loop is, and what to do next
product.status - Update the product details for this workspace
product.update