Menu
Synthesize user interviews, survey responses, or forum threads from genealogists into structured product insights and design prompts.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Audit place resolution against a real database to find outliers — places that resolve to the wrong country, state, or sub-region. Use when investigating bad map pins, validating new gazetteers, or debugging "why did Richmond, Kalifornien USA land in Canada?". Pairs with the `gazetteers` skill (build/extend) — this skill is for *testing* placement on real data.
Use whenever dispatching subagents in this project — implementer, spec-reviewer, or code-quality-reviewer. Wraps the upstream `superpowers:subagent-driven-development` skill with project-local prompt templates that center user goals (not just spec compliance) and require the dispatcher to verify user-observable outcomes before marking work done.
Decide which Playwright tier to run, capture the right evidence for plan close-out, and route work into the right project. Use when finishing a plan (gate before archive), when adding an e2e test (which project gets the spec), when CI fails on `[boot]` / `[crud]` / `[panels]` / `[reactivity]` / `[imports]` and you need to know which user-goal it protects, or when the question is "is `npm run test:e2e` enough or do I need `:full`?".
Extend the Släktforskning MCP server — add new tools to createProdServer.ts / createDevServer.ts, wire window.api bindings in tauri-window-api.ts + static-api stubs, test via tests/unit/mcp.test.ts, debug agent-facing IPC. Also covers the dev MCP HTTP bridge (src-tauri/src/ui_server.rs) used by ui_screenshot / ui_click / ui_eval / chart_*. Use when modifying anything in src/mcp/, src/renderer/tauri-window-api.ts, or src-tauri/src/ui_server.rs. Distinct from `slaktforskning-mcp` (which is for an agent *using* the MCP tools to do genealogy work for the user).
Build, extend, and debug gazetteers for place resolution. Use when adding new country gazetteers, modifying build scripts, debugging place matching, or questions about the gazetteer system (types, resolver, normalization, data sources).
Use when actually doing genealogy research with the Släktforskning MCP — building real family trees end-to-end, exporting/round-tripping, attaching media + face tags, importing custom gazetteers for places not in the bundled set. Pairs with `slaktforskning-mcp` (which lists the tools). Distinct from `slaktforskning-mcp-dev` (extending the MCP server itself). Triggers on real research workflows, "research X family", "test the app end-to-end", "set up a Bernadotte-style demo".
| name | interview-synthesis |
| description | Synthesize user interviews, survey responses, or forum threads from genealogists into structured product insights and design prompts. |
This skill turns raw qualitative research — interview transcripts, notes, survey answers, forum threads — into structured product insights that drive genealogy app ideation.
When reading interview material, extract insights across these dimensions:
What frustrates users? Where do they get stuck? What takes too long? What feels broken or missing?
How do users actually work? What's their step-by-step process for:
How do users think about genealogy data? What metaphors do they use? How do they conceptualize relationships, time, and evidence? Mismatches between user mental models and software models are fertile ground for design improvement.
What words do genealogists use that software doesn't reflect? (e.g., "my brick wall", "collateral lines", "cluster research", "reasonably exhaustive search")
What are users ultimately trying to accomplish? (e.g., "prove my grandfather's immigration story", "connect with living relatives", "write a family history book")
What do users love? What makes them feel accomplished or excited?
What tools or hacks do users employ because the software doesn't support their need? Workarounds reveal strong unmet needs.
Numbered list of 5-10 themes, each with:
Bulleted "How might we..." questions derived from the themes. These are design prompts. Example: "How might we help users track which sources they've already searched, not just what they found?"
If the material reveals distinct user types, sketch 2-3 personas with: name/label, goals, frustrations, technical comfort, how they use genealogy apps.
Top 5 product directions suggested by the research.
Släktforskning is a local-first, privacy-preserving desktop genealogy app with a built-in MCP server for AI-assisted research. When synthesizing interviews, pay special attention to:
These align with Släktforskning's core differentiators. See docs/PLAN.md for the current roadmap to connect insights to planned features.
