Once the Player Hands Over Information, Who in the World Should Know?

After the player hands a survey report to the workshop, first update what the people who actually read it know, then spread that information through specific events. Do not turn “handed it to the workshop” directly into “the workshop and the council both know.” Each transmission should answer at least these questions: who received which parts, from whom, when, and by what means?

Keep three things separate: permission to receive, actual knowledge, and permission to pass information on. Someone may have permission to read a document but not yet have opened its envelope; someone else may know a secret after an unauthorized peek. Permissions determine whether an action crosses a boundary, while transmission records determine what a character knows. Knowing what a report says is also different from believing it.

The following “Mistbridge Survey Room” is a fictional teaching example. Its characters, report, and events are original illustrations, not a real user case, empirical test results, or DramaFork product features. The player discovers a discrepancy in Mistbridge’s measurements and chooses to submit the report to the workshop for verification first, rather than to the council to initiate public questioning.

Split the report into three parts: A is “the eastern scale has an error,” B is “raw readings and measurement-point locations,” and C is “the signature of Qinghe, the recorder.” The repairer needs A, the verifier needs B, and the council could use C to find Qinghe and ask further questions. The signature identifies only the recorder; it does not establish who delivered the report. In this example, the player delivered it.

An agreement describes what may happen; a record describes what has happened. After a leak, append the facts without changing the original restriction into permission.

“The workshop” cannot stand in for a recipient list. If Cen Duo has accepted the envelope but not opened it, he knows only that the player delivered a report; he cannot immediately mention Qinghe. Add a record of knowledge only when there is evidence that someone read or clearly heard the content, recording authorization separately. An unauthorized peek can produce knowledge; meeting the authorization conditions without reading does not.

Reusable fields include event ID, time, sender, recipient, content and version, channel, authorization status, evidence of receipt, and state of understanding. Evidence might be a restatement, annotations, or a scene explicitly showing someone reading the document. If only the act of sending is established, record “sent; whether it was read is unknown.”

There is no need to record every casual conversation. Record only transmissions that change later dialogue, actions, or responsibility.

Ajian needs to check the scale and temporarily leaves the sheet containing B on a shared tracing table. When the council clerk comes to collect old drawings, the clerk sees and copies one line: “Eastern measurement point No. 2: reading twelve.” This line includes no normal range, assessment of error, verification status, or signature.

The scene conditions for the leak are that the sheet has not been put away, the clerk enters the scene, and the readings face upward. The scene must also explicitly establish that the clerk actually reads it. Having entered the workshop is not enough to establish a leak.

Append event four: “The supplementary sheet containing B, in Ajian’s care, is exposed on the tracing table. The clerk reads and copies the line above on the spot, without authorization, learning only the measurement point’s name and its recorded reading.” Ajian is not deliberately informing on anyone, but the way the sheet is handled forms part of the transmission path.

In the next scene, the clerk can say, “The reading recorded for eastern measurement point No. 2 is twelve. What does this line refer to?” The clerk cannot simply say, “Why is the reading abnormal at that eastern point you’re verifying?” The clerk has neither obtained the verification status nor acquired the basis needed to judge the reading abnormal. If the line is meant as a guess or a probe, label it as such rather than treating it as established knowledge.

Nor can the clerk conclude, “Qinghe handed you the complete report.” The clerk has not read C and does not know the full report’s contents or how the handoff occurred. Moreover, the player delivered the report in this example; Qinghe is the recorder whose signature appears on it. Even reading the full original later would not establish from the signature alone that Qinghe delivered it personally. How the handoff occurred needs an independent source.

If the clerk deliberately claims, “I’ve read the complete report,” record that as a lie intended to probe for information. If Cen Duo is bluffed into revealing more, add a new verbal transmission. Do not use the lie to infer that the clerk already knew the information.

The player should not gain unexplained access to the author’s complete record either. You could first let the player see the clerk tuck the copied sheet into the old drawings, then hear the clerk quote that line, giving the player a reason to investigate. These clues still do not prove who deliberately leaked anything. Maintain the path known to the author separately from the clues available to the player.

The same task can offer three ways of handing over the report without making secrecy the only correct answer.

Hand it directly to the council, and the clerk reads the complete original in person. The player gains a route to formal questioning, and Qinghe’s signature also comes to the clerk’s attention. Other council members do not learn everything simultaneously. The clerk knows the player delivered it because of the in-person handoff, not the signature.

Give only the workshop a copy with the signature removed, and Ajian can still check B. The tradeoff is that if questions about the handwriting arise, the workshop cannot use this copy alone to locate the recorder; the player must provide more information. You cannot both hide the source and have the verifier find that source without any other clues.

Allow the workshop to pass A to the council after new measurements confirm it, and you add conditional authorization. Meeting the condition means only that sending is permitted. Cen Duo must still write a note, and the clerk must receive and read it before new knowledge is established.

If the player revokes permission before sending, record when that happens. Even if something has been sent but not read, that alone does not establish knowledge. If the note has already been read, revoking permission cannot erase knowledge; it can only restrict subsequent sharing within the scope of that authorization.

These differences arise from content, timing, and recipients. There is no need to invent extra trust scores. By choosing how widely information spreads, the player also chooses who will ask questions later and who will need to come back for additional evidence.

Take the three most important lines in the next scene and trace each one back: through which event did the speaker learn this, did they receive the original, a summary, or a rumor, and are they treating inference as fact? Names, numbers, verification status, judgments of abnormality, and claims about who delivered the report each need their own basis. If no source exists, add a transmission scene or narrow the line.

The first counterexample is “They belong to the same workshop, so they know whose signature is on the report.” Organizational ties are not a reading event. Tao Qi can arrange repairs using only A unless someone separately tells him C. Even knowing C does not automatically reveal who delivered the report.

The second is “They received the supplementary sheet, so they believe the bridge has a problem.” Ajian can see B and still suspect that the instrument is out of calibration. Record knowledge and belief separately so arguments can center on evidence.

The third is “The report was taken back, so the secret is safe again.” Taking back the paper changes who holds it; it does not erase memories or copies. Record a corrected version and its recipients separately. Someone who has not received the correction may still quote the old reading.

This method suits the directed flow of a small number of key pieces of information. A reading in a public square can be recorded as a group event for people confirmed to have been present and to have clearly heard the content, but it does not establish that the entire city knows. Retain every person for whom there is evidence of receipt, the version they learned, and subsequent corrections. Do not overwrite earlier readers with the latest recipient. Check what characters can state with certainty item by item against the content they actually received and independent event evidence.