# Puzzle Kits - Open Questions for Ralph Everything a pupil or teacher reads that is NOT canon-verbatim is listed here for your review, plus every judgment call I made. Nothing here blocks use of the kits; edits are one-line changes in tools/puzzle_gen.py (INSTRUCTIONS dict) or tools/render_kit.py, then `python3 tools/build_all.py` regenerates everything. ## 1. Instructional / functional text inventory (needs your sign-off) From tools/puzzle_gen.py INSTRUCTIONS (appears on puzzle pages): - Kit header: "TOTAL MAYHEM PUZZLE KIT" (+ " - BOOK N" / theme label); cover masthead renders it as "TOTAL MAYHEM" over letterspaced "P U Z Z L E K I T"; cover badge words "BOOK N" / "THEME KIT" - Cover contents label: "IN THIS KIT" - Cover note: "Every name, fact and quote in this kit is taken word-for-word from the Total Mayhem canon records. The answer key cites a source for every answer (record id + book, chapter, page)." - Theme-kit source line variant: "Source: canon records in canon/data/ (cited per answer)" - Word search title "WORD SEARCH"; instruction "Find each name in the grid. Names run in straight lines - forward, backward, up, down, or diagonally. Ignore spaces, hyphens, apostrophes and accents." - Match-up title "MATCH-UP"; instruction "Write the letter of the matching description next to each number." - Theme variants: "Match each Move number to its name." / "Match each invention to its inventor." - Fill-in title "FILL IN THE BLANKS"; instruction "Complete each line from the book using the word bank. Each answer is used exactly once."; box label "WORD BANK" - Codes title "CODE-BREAKING"; example box label "FROM THE BOOKS"; "Decode each name. Slashes separate the letters."; "Each item spells a name with the NATO alphabet. Write the name." - Table labels: "MORSE CODE - the system Dash and Rob use" / "NATO PHONETIC ALPHABET - as printed in the Week One Almanac" - Flash title "KB-15 FLASH CODES"; instruction "Match each KB-15 flash pattern to what it signals." - Answer key title "ANSWER KEY"; note "Citations show the canon record id, then book, chapter and page. Raw refs use book:chapter:page." - Footer tag "Total Mayhem Puzzle Kit" Chrome in tools/render_kit.py: - "Name" write-in line (top right of puzzle pages); "page N of M" footers - Cover source line: "Source: canon records for in canon/data/" - Answer-key labels: "Morse 1.", "NATO 1.", "Flash 1 - B", "1 - C", word-search locators "r4 c5 down-right" and direction words (right/left/up/down/down-right/up-right/down-left/up-left) - Morse table footnote "canon anchors: f_b01_011, f_al1_300" - Citation format: "record_id; Book N, ch. C, p. P [bNN:cCC:pPPP]" Q: Approve as-is, or edit wording? (All are deliberately flat/functional - nothing in book voice, per the ground rules.) Visual brand elements (drawn shapes, no words, added on Ralph's "less municipal" request, then revised print-friendly on his follow-up - no large solid fills anywhere): double-frame cover masthead with big black type, line-drawn hazard-stripe bands, a double-outline starburst cover badge, outlined label tabs on boxes, square-bullet titles over heavy rules, and a dots-and-dashes cover motif drawing Dash's printed Morse code name - that motif is a cited canon item (f_b01_011 substring, cover.morse_strip in each content JSON, cite printed beneath it; the dots/dashes are the only filled shapes and they are small). Q: keep, tune, or drop any of these? ## 2. Conventions I adopted (mechanical, all machine-verified) - BLANKING: fill-in items blank exactly one token of a canon quote; the blank is drawn with one underscore per letter (kids get a length hint). OK? - SUBSTRING: KB-15 flash lines are split at the printed colon into pattern/meaning; Move entities split into "Move #N" / name; covers for Books 1-7 quote the per-book line of the almanac's coverage list (f_al1_399, e.g. "Monday Into the Cave of Thieves"). OK? - LETTERIZATION: word-search grids use the name uppercased with non-letters dropped and accents mapped to base letters (Footmøld -> FOOTMOLD in the grid; the printed word list always shows the canon spelling). OK? - ENCODING: Morse items are canon names/aliases encoded with the standard international table; the books themselves use it (Dash's and Rob's Morse code names decode with exactly this table), and those two facts are cited on every codes page. The full A-Z Morse table printed as a decoder is real-world reference material, NOT book content - flagging explicitly. The NATO table needs no such caveat: it is quoted verbatim from canon (f_al1_250, Week One Almanac p. 209, with your F=Foxtrot ruling). OK? - Word-search filler letters are random, seeded per kit, drawn from the answer words' letter distribution; the auditor proves each answer appears exactly once. OK? ## 3. Judgment calls to confirm 1. Books 8-14 have no canon title record, so their covers say only "Book 8" etc. Options: leave as-is; you supply editorial titles; or a future canon pass adds title facts (the in-book ads/blurbs may support this). 2. Book 12 is ONE kit merging volumes b12a + b12b (one story; citations keep the volume-specific refs). Right call? 3. Morse items may be multi-word names (e.g. "Rob Newman" -> ".-./---/-.../-././.--/--/.-/-." with no word gap; the instruction says to ignore spaces). Alternative: restrict Morse items to single-word names. 4. Matching descriptions are canon description_as_written and sometimes name OTHER characters ("Dash's best friend" for Rob Newman) - inherent to the source text; the auditor only filters descriptions containing the entity's OWN name. Fine? 5. Fill-in blank words are picked mechanically (capitalized > numeric > longest token, unique within the puzzle). Semantic guessability is not machine-checkable - worth a human spot-check of one or two kits. 6. Word searches allow all 8 directions (grades ~3+). Want an easier 4-direction / no-reverse variant for younger readers? One-line change. 7. Theme kits: I built two ("Moves and Scallywags", "Gadgets and Codes"). Obvious candidates for more: Swedhump Elementary (people and places), Food (34 food entities), Creatures. Want them? 8. The almanac has no standalone book kit (its content powers the theme kits and the NATO table). Want a dedicated Almanac kit? 9. Dashes: canon-verbatim strings keep their printed en/em-dashes (e.g. "Red – on-off 1 second intervals continuous"); all text I authored uses hyphens only, per the standing rule. Confirm this split is as intended. 10. The kits' fact quotes occasionally embed printed misprints kept as printed in canon (e.g. spellings the books themselves vary). They will surface verbatim in puzzles. That is the point of canon-verbatim, but flagging so a teacher-facing note isn't a surprise request later. ## 4. Not done (scope guards) - No illustrations/assets (asset registry is a future canon phase). - No crosswords (clue text would require authored prose - excluded by rule). - No AI-written story text anywhere; quiz formats that need invented wrong answers (true/false, distractor multiple choice) were deliberately avoided - the word bank IS the answer set.