# Homeric Product Discovery

Continuous-discovery coaching skills (Teresa Torres's Continuous Discovery Habits + Strategyzer experiment practice) plus the free Product OKR skills, callable from Claude, Claude Code, Cursor, Windsurf, and Atlassian Rovo over the Model Context Protocol.

MCP endpoint: https://mcp.homeric.ai/product-discovery

## Connect your agent

Add https://mcp.homeric.ai/product-discovery as a remote MCP server (HTTP transport). On the first tool call your agent opens a browser to sign in with your Homeric account — the server is an OAuth resource server and WorkOS AuthKit issues the tokens.

Example mcpServers config (Claude Code, Cursor, Windsurf):

    {
      "mcpServers": {
        "product-discovery": {
          "type": "http",
          "url": "https://mcp.homeric.ai/product-discovery"
        }
      }
    }

## Skills (12)

### discovery_create_experiments
Create an Experiment — requires the Product Discovery subscription

Turns one assumption into a runnable Experiment Card per Strategyzer's Testing Business Ideas: 1-3 recommended types from the 44-experiment library, the four statements on their stems, risk type, lifecycle fields, tree connections, and the artifact the test needs.

### discovery_review_experiments
Review an Experiment — requires the Product Discovery subscription

The deep coaching read of an Experiment Card: a markdown narrative walking hypothesis quality, experiment fit, evidence discipline, criteria rigor — plus learnings when completed — with the why and framework references. Use experiment-score for the fast scorecard.

### discovery_score_experiments
Score an Experiment — requires the Product Discovery subscription

A fast Experiment Card scorecard: star ratings for Hypothesis, Test, Metric, Criteria — plus Learnings when the card is completed — with a couple of coaching nudges to improve it. Scored against Strategyzer's Testing Business Ideas. Use experiment-review for the deep read.

### homeric_message_product_coach
Message Product Coach

Message a human Homeric product coach — ask a question, or check for a reply to a previous one.

### discovery_create_opportunity_trees
Create an Opportunity Solution Tree — requires the Product Discovery subscription

Builds a Homeric Opportunity Solution Tree from a Key Result plus customer evidence, per Teresa Torres: HMW-framed problems traced to evidence, one target problem, 2-3 concise ideas, and experiment stubs for the riskiest idea. Siblings: opportunity-tree-score / review.

### discovery_review_opportunity_trees
Review an Opportunity Solution Tree — requires the Product Discovery subscription

The deep coaching read of an Opportunity Solution Tree on the Homeric mapping: a markdown narrative walking outcome quality, problem framing (HMW), idea quality, and experiment linkage — with the why and framework references. Use opportunity-tree-score for the fast scorecard.

### discovery_score_opportunity_trees
Score an Opportunity Solution Tree — requires the Product Discovery subscription

A fast OST scorecard: star ratings for Outcome, Problems, Ideas, and Experiments — the tree's four layers on the Homeric mapping (KR root, HMW problems, concise ideas, experiment stubs) — with a couple of coaching nudges. Use opportunity-tree-review for the deep read.

### discovery_sync_ost_to_jira
Sync an Opportunity Tree to Jira — requires the Product Discovery subscription

Pushes opportunity-tree work into Jira Product Discovery and Atlassian Goals via the user's Atlassian MCP connection — problems, ideas, experiments, the Objective's Goal — wiring the links and contributions that make JPD's tree view render. Composable actions.

### discovery_create_product_okrs
Create Product OKRs

Proposes an improved OKR set on top of the user's input (or drafts one from a strategic priority brief) with per-dimension rationale per the SVPG/Wodtke framework. Use `product-okr-score` for a quick score; use `product-okr-review` for a thorough read without a rewrite.

### discovery_review_product_okrs
Review Product OKRs

Conversational coaching review of a product OKR set against the SVPG/Wodtke framework. Calls out what's working and what to address per dimension, in a warm teammate voice. Use `product-okr-score` for a quick verdict; use `product-okr-create` for a rewrite.

### discovery_score_product_okrs
Score Product OKRs

A fast OKR scorecard: a 3-star rating across Context, Objective, Key Results, and Ops Metrics — scored underneath against the full SVPG/Wodtke framework — plus a few friendly ideas to improve. Use product-okr-review for the deep per-dimension read.

### homeric_request_skills
Request a Homeric Skill

Capture a skill the user wishes Homeric offered. First checks whether a catalog skill already covers the need and points there; otherwise structures the unmet need into a clean request and routes it to the human Homeric coaches.

---
Manage your skills, connections, and activity at https://mcp.homeric.ai