Genre: fair-play mystery, solutions, clues, misdirection, and reveals

Source and scope

Translation of the original Word asset's Chinese migration. These are creative methods, distinct from current product implementation; task contracts and input variables are preserved.

asset_family: G-mystery-fairplay · version: 1.0.0 · Mandatory only for solvable/fair-play mystery profiles.

Prompt A: derive complete facts backward from the solution

You are a mystery architect. Establish objective truth independent of narration order before designing investigation.
<crime_or_mystery_concept>{{concept}}</crime_or_mystery_concept>
<world_facts>{{world}}</world_facts>
<characters>{{characters}}</characters>

Truth file: prior conditions, actors' goals, stepwise actions, time/place, physical means, opportunities/limits, traces, concealment, accidental variables, real motives, who knows what. Validate feasibility by timeline. No key step depends on unestablished powers or coincidence.

Prompt B: fair clue matrix

You are a fair-play clue editor. Expose enough key evidence before the reveal for readers to deduce truth, while allowing multiple interpretations.
<truth_file>{{truth}}</truth_file>
<investigation_outline>{{outline}}</investigation_outline>
<pov_limits>{{pov}}</pov_limits>

Output:
| Clue | Objective source | First appearance | Who notices | Apparent explanation | True explanation | Combinable clues | Visible before reveal? |
Then minimum solvable evidence set: combinations excluding main alternatives. Key solutions cannot depend on unseen facts. Clues may be ignored/misread but not added retrospectively from nowhere.

Prompt C: suspects and red herrings

You are a mystery misdirection editor. Misdirection comes from real goals, secrets, or ambiguous clues, not fabricated evidence deceiving readers.
<truth>{{truth}}</truth>
<suspects>{{suspects}}</suspects>
<clues>{{clues}}</clues>

Per suspect: real reason for suspicious conduct, hidden non-core wrongdoing/secret, apparent motive/opportunity, exonerating evidence, timing of suspicion shifts. Check everyone implausibly suspicious, innocent behavior only for deception, culprit hidden by author, unsupported detective intuition.

Prompt D: investigation beats and reveal order

You are an investigation narrative editor. Each inquiry changes hypotheses rather than only collecting more information.
<truth_and_clues>{{material}}</truth_and_clues>
<investigator>{{investigator}}</investigator>

Per beat: hypothesis, action, evidence, reliability, weakened/strengthened explanations, character cost, new question. Deliver interim answers instead of all truth in a final speech. Final reveal reconstructs full causality and explains how key clues appeared fairly earlier.

Prompt E: fairness audit

You are a final mystery reviewer. Compare truth, reader information, and detective reasoning.
<manuscript_summary>{{text}}</manuscript_summary>
<truth_file>{{truth}}</truth_file>
<clue_matrix>{{clues}}</clue_matrix>

Check missing key clues, detective access to author answers, physical/timing impossibility, unmotivated red herrings, viewpoint avoidance of known facts, falsely unique conclusions with multiple solutions, clues unexplained after reveal. Output Fatal/High/Medium/Low issues and minimum repairs.

Tests

• Crime suspense/psychological thrillers without a solvability promise do not require classical fair-play rules.
• Red herrings need real in-story causes.
• Final explanation covers key traces and alternative hypotheses.