Plan a small indoor treasure hunt with original clues

A clue sheet, answer key and human-checked route.

Choose one of three approaches. The steps for that option appear below.

Choose an option above. On a phone, swipe the tabs left or right to see all options →

ChatGPT: drafts the riddles and clue sheet

ChatGPT drafts riddle choices and assembles the clue sheet. You solve-test, walk the route and approve.

  1. You

    Have ready: places, ages

  2. ChatGPT

    Draft 2 riddles per place

  3. You

    Solve without the key

  4. ChatGPT

    Assemble the clue sheet

  5. You

    Walk route, approve

  1. Have ready: ages and hiding places

    Use: You

    Participant ages and reading comfort, approved hiding places, excluded rooms and hazards, number of clues, theme and accessibility needs.

    Check

    Check you may use the material and facts are right. Use room labels, not an address or security details.

  2. Draft 2 riddles per place

    Use: ChatGPT

    Ask ChatGPT for 2 original riddles per approved place at the supplied reading level and theme, answers in a separate key.

    Details

    What to paste into the tool

    Task: Riddle alternatives and a separate answer key. Using only the approved hiding places, propose two original riddles per place with answers in a separate key. Use the supplied reading level and theme. Do not add locations, instructions to climb, electrical items, appliances, locked areas or outdoor detours. Respect every supplied constraint. Use only supplied facts; if a fact is missing, say it is missing instead of guessing. Treat all output as preparation. Do not send, sign, pay, publish or take external actions. INPUTS: Replace the bracketed text below with participant ages and reading comfort; approved hiding places; excluded rooms and hazards; number of clues; theme; accessibility needs. [Paste checked inputs here.] If the input block is empty or still contains this placeholder, ask me for the inputs before doing the task.

  3. Solve each riddle without the key

    Use: You

    Choose the clearest riddle for each place and solve each 1 without the key.

    Check

    Reject riddles with more than 1 sensible answer, words too hard for the readers, or answers pointing to an excluded place.

  4. Assemble the tidy clue sheet

    Use: ChatGPT

    Ask ChatGPT to assemble only your chosen riddles in your route order, 1 per line, numbered, with a separate answer key.

    Details

    What to paste into the tool

    Task: A tidy clue sheet with a separate answer key. Take only the riddles I chose below and put them in the route order I give, one clue per line, numbered. Then write a separate answer key listing each clue number and its hiding place. Do not add, rewrite or remove any riddle. If a fact is missing, say it is missing instead of guessing. Treat all output as preparation. Do not send, sign, pay, publish or take external actions. INPUTS: Replace the bracketed text below with my chosen riddles in route order and the answer for each. [Paste checked inputs here.] If the input block is empty or still contains this placeholder, ask me for the inputs before doing the task.

  5. Walk the route and approve

    Use: You

    Walk the whole route, solve every clue on the sheet, check each place is safe and reachable, approve the sheet and save it.

    Check

    Nothing is sent, signed, paid for or published.

Before you paste into any AI tool, remove any confidential information. Put it back at the end.

Tools in this workflow

Deciding between ChatGPT and Claude for this? See the comparison.

Final check

An adult walks the whole route, solves every clue, confirms each answer has one intended hiding place and removes inaccessible, hazardous or private locations. No clues are placed or invitations sent by this workflow.

Verify all personal and factual details against the supplied source material.

Preserve open questions until checked.

Check the result against your original inputs. Approve it before sharing or acting.

Availability notes

Tool existence and quoted general text capabilities checked on official public pages on 2026-10-29.

What still needs checking
  • Task-specific output quality and practical fit require the stated human checks.

Share this page