SyllvanicDevelopments · Facilitator
Prompt suggestions
Add one or more of these plain-language requirements after the Facilitator instructions to control what the questionnaire asks and how it is organized.
Counts, grouping, and coverage
- Exact question count
Create exactly 30 questions.- Maximum question count
Use no more than 20 questions; use fewer if they fully cover the subject.- Exact subject count
Cover exactly 150 unique anime titles across the complete questionnaire.- Fixed grouping
Place exactly five anime titles in each question.- No duplicates
Do not repeat a title, option, or subject in more than one question.- Category allocation
Use exactly five questions for each of these categories: scope, preferences, constraints, risks, workflow, and success criteria.
Question and answer behavior
- Choose one
Use multiple-choice only when exactly one option should be selected.- Choose all that apply
Use multiple-select because I may select more than one answer in each question.- Optional explanation
Use hybrid when I should select one option and may optionally explain it.- Open response
Use pure-text when predefined choices would constrain the answer unhelpfully.- Ranking
Use pure-text and ask me to enter the items in ranked order; do not simulate ranking with multiple-select.- Required versus optional
Make only questions essential to the stated outcome required. Mark exploratory questions optional.
Truthful exits and participant agency
Choose an exit that describes why the substantive answers may not be truthful. These
are ordinary options in multiple-choice and hybrid questions. Only multiple-select
exits belong in exclusiveOptions.
- Premise does not apply
Include “Not applicable” when the question may not apply to the participant.- Information is unavailable
Include “I don’t know” or “Not enough information” when uncertainty is legitimate.- Sensitive disclosure
For sensitive required questions, include “Prefer not to answer” as an ordinary option.- No preference
For preference questions, include “No preference” when neutrality is legitimate.- Situation varies
For preferences that may change by context, include “Depends on the project”.- Safe refusal
For authorization or safety decisions, include an explicit choice such as “Do not proceed”.- Unlisted answer
Use “Other” when another response is possible; prefer hybrid when an optional explanation would help.- Multiple-select exit
Add “None of the above” as the final option and list that exact label in exclusiveOptions.
Question quality and output discipline
- Decision relevance
Ask only questions whose answers could change the recommendation, implementation, or risk assessment.- Neutral wording
Use neutral, non-leading wording and do not imply that one response is preferred.- One dimension at a time
Each question must measure one distinct issue; do not ask the same underlying question twice.- Balanced options
Make options mutually distinct and comparable at approximately the same level of detail.- Consistent scales
Arrange scale options consistently from least to most without reversing direction between questions.- Stable identifiers
Use sequential IDs Q01 through Q30 with no gaps or duplicates.- Final verification
Before responding, verify the question count, unique-item count, option count, IDs, question types, and exclusiveOptions references.
Complete examples
Anime-title coverage
In exactly 30 questions, determine which of exactly 150 unique anime titles I have seen. Use five distinct anime titles per question, with no title repeated anywhere in the questionnaire. Make every question multiple-select and required. Add “None of the above” as the sixth and final option in every question, and list it exactly in exclusiveOptions. Do not count “None of the above” among the 150 titles.
Requirements discovery
In no more than 15 questions, identify the requirements needed to design this website. Cover audience, primary task, content, visual direction, accessibility, privacy, integrations, schedule, and success criteria. Use multiple-choice only for genuinely exclusive decisions, multiple-select for compatible requirements, hybrid when an optional explanation would help, and pure-text when choices would not be faithful. Make only decisions that block implementation required.
Sensitive collaboration boundaries
Ask up to 12 neutral questions about collaboration boundaries. Every required multiple-choice or hybrid question involving privacy, attribution, identity, ownership, or personal disclosure must include a fitting option such as “Prefer not to answer”, “Not decided yet”, or “Requires a separate agreement”. Do not use “None of the above” when the real issue is declining disclosure.
Prioritization
Ask me to prioritize ten features. First use one multiple-select question to identify every feature I consider important. Then use pure-text questions asking me to list the selected features in priority order and explain hard dependencies. Do not represent ranking as a multiple-select answer.
Decision-ready questionnaire
Ask exactly 10 questions needed to choose among the three proposed approaches. Avoid background questions that will not affect the decision. Ensure that each response could change the recommendation, implementation, or risk assessment. Include “Not enough information” when the decision cannot be made responsibly from the listed choices.
Reusable finishing instruction
Make the package decision-useful rather than merely comprehensive. Ask each underlying question once, use neutral wording, keep options distinct, and make only blocking questions required. Give required choice questions a context-appropriate truthful exit when the listed substantive answers may not apply, information may be unavailable, or disclosure may be declined. Verify counts, uniqueness, option limits, IDs, question-type semantics, and exclusiveOptions before returning JSON only.