01 — the problem
🔇

Agents receive vague input

"Build me a dashboard" misses context, constraints, and tradeoffs. The agent guesses. You get the wrong thing. Repeat.

02 — the solution

questy asks the right questions

A structured Q&A skill that reads your project docs, detects gaps and conflicts, then generates targeted questions in typed batches.

standards/ plans/ questions pendings
03 — the type system

6 Thinking Hats → the types

Every question wears exactly one hat; the hat is the type (what it's about AND how to think about it). No catalogue — hats replace the legacy vision/arch/trust/data/integration.

⬜ facts 🟥 heart ⬛ risk 🟨 optimism 🟩 creative 🟦 make
docs→🟨 design→🟩 threats→⬛ ux→🟥 integration→🟦 data→⬜
04 — enrichment addons

Beyond the core skill

Optional modules that extend questy with structure, automation, and integrations.

TreeCreate questy-docs/, plans/, standards/
GitSynced detection, branch rules
Agents.mdPer-project tuning
HTMLInteractive form via questy-html
GitHubWorkflows, PR standards
ArchiveResolved → archive.md
ProArchitecture decision records
Scripts12 helper scripts
WorkflowsSynced, merge rules, blueprint
05 — how it's done

Start to finish

flove.org/development/standards/skills/questy/
Generate a structured prompt for questy — the skill that reads your project docs, detects gaps, and asks the right questions.
React AppNode.js APIFlove AppCLI Tool
Createproject name...docsok
flove.org/development/standards/skills/questy/
Thinking Hats the type
facts 1heart 1risk 1optimism 0creative 0make 0
tones · focus
flove.org/development/standards/skills/questy/
QuestionsAddons
flove.org/development/standards/skills/questy/
# questy prompt
project: my-app · mode: numbered+explained
hats: facts, heart, risk (×2), creative
addons: tree, git, scripts, workflows
Copy Prompt ✓
→ paste into your agent to start the Q&A
Stalinks · flove.org/docs · flove.org