Does this template give an AI automatic memory?
No. It is a manual text record you review and provide yourself. It makes no promise that a tool will remember, load, or enforce it.
Free manual template
Carry approved story decisions into your next conversation without turning discarded branches or possible future beats into canon. Copy the complete template, compare the original example with its sources, and check each claim yourself.
Keep your original story and source excerpts. This manual handover records decisions, corrections, unknowns, and what each character knows at a stated chapter boundary. It does not make a writing tool remember or enforce those decisions.
The template is free to copy without an account. thredly summaries are an optional separate workflow; this example is not generated product output. For tool-specific context, see NovelAI chat memory and saving NovelAI chats.
Copy this, then replace the prompts with the details from your conversation.
Free manual fiction canon handover template Copy this into a plain text document. Keep the source excerpts beside it. Replace the brackets yourself and review each claim before reusing the handover in a new conversation. This template adds a decision ledger, branch status, and chapter-bounded character knowledge to whatever method you use to save your writing. It is a manual record; it makes no promise about a tool remembering, loading, or enforcing it. Project and evidence boundary Title: [title] Current draft or version: [version, approval source] Material actually supplied for this handover: [turns, excerpts, chapters] Material only reported as available: [reported scope and source] Missing material: [what is absent; do not claim to have read it] Next permitted task: [task and source] Tasks to hold for author approval: [drafting, plot resolution, other] Current canon - [ID] [single claim] | status: author approved | chapter or time boundary: [boundary] | source: [ID plus exact quote] | approval source: [ID] Corrections and supersession - [old ID and claim] | earlier status: [approved/proposed] | now: [superseded/rejected] | changed by: [ID plus exact quote] | replacement: [new ID or none] - [why this is a replacement, not an additional event or alternate timeline] Branches and proposed beats - [branch or proposal ID] [idea] | status: [proposed/rejected/provisional] | source: [ID plus exact quote] | author approval: [absent/source] | occurred in current canon: [yes/no/unknown] - Do not convert a suggestion into canon because it is vivid, recent, or convenient. Story chronology - [event] | chapter or story time: [known boundary] | state: [established/planned/proposed] | source: [ID plus exact quote] | unknown timing: [date, placement, or relative order not established] - A weekday is not a calendar date or a chapter placement. Do not infer event order from source-turn order. Character knowledge at the stopping point - [character] | as of [chapter/scene] | has seen/knows/believes: [supported statement] | does not know or has not seen: [supported statement] | unknown: [unspecified details] | source: [ID plus exact quote] - Separate seeing a clue from understanding it. Separate a possible later discovery from present knowledge. Unresolved questions - [question] | state: undecided | last explicit author statement: [ID plus exact quote] | no inference allowed from: [tempting but insufficient clue] Review checklist [ ] Every current claim has a source and explicit authority. [ ] Every correction links the old claim to the replacement or rejection. [ ] Proposed, provisional, rejected, and superseded material stays visible and labeled. [ ] Character knowledge has a chapter or scene cutoff. [ ] Unknowns have not been filled with guesses. [ ] Chronology distinguishes established chapter order from unknown event timing. [ ] Reported file availability is separate from material actually supplied or inspected. [ ] The next task respects the author's boundary. Checked by: [author or collaborator] Review date: [date] Author decision: [needs correction/approved for this task] Questions for the author: [questions or none] Reusable review prompt Read the supplied source excerpts and this handover. Check each claim against its source ID and exact quote. Treat explicit author decisions and corrections as authority; do not assume that the latest suggestion is approved. List contradictions, unsupported claims, stale decisions, branch leakage, and character-knowledge errors with source IDs. If a check requires missing chapter prose, say what prose is needed. Keep any optional new beats labeled proposed, preserve unresolved questions, and stay within the stated next task. Ask the author to approve any proposed canon change before using it as established fact.
F01 (invented author): The working title is The Bell at Greyharbor. In an early brainstorm, branch A would have Mara steal a brass key from Ivo in chapter 3. That is an idea to test, not approved canon. The missing bell is the mystery; I have not decided who stole it.
F02 (invented collaborator): Here are three possibilities, all proposals: the bell disappears on Tuesday; the brass key opens the lighthouse door; Ivo notices a concealed mark in chapter 4. We could use one, several, or none of these.
F03 (invented author): For the current planning pass, approve Tuesday as the day the bell was stolen. Keep the other F02 ideas as proposals. I have not approved branch A or a thief identity.
F04 (invented author): Reject branch A: Mara did not steal the brass key from Ivo. In canon, Ivo willingly gives Mara the brass key in chapter 3. Keep the rejected theft idea in the history so it does not drift back into a summary.
F05 (invented author): The brass key opens the archive cabinet. Reject the proposal that it opens the lighthouse door. Do not imply that Mara can use this key to enter the lighthouse.
F06 (invented author): In chapter 4, Mara sees a concealed mark on the inside of the archive cabinet. At the end of chapter 4, Ivo has not seen or discovered that mark. Reject the F02 proposal that Ivo notices it in chapter 4. Seeing the mark does not establish its meaning; its meaning is still undecided.
F07 (invented author): Correction to F03: the bell was stolen on Thursday, not Tuesday. Thursday replaces Tuesday in current canon. Keep Tuesday only as a superseded decision, not as a second theft or a separate timeline.
F08 (invented collaborator): A possible chapter 5 beat: Ivo discovers the concealed mark when Mara shows him the cabinet. This is a future-beat proposal, not an event that has happened or an author-approved change. Another possible beat is Mara deciding whether to tell him before opening the cabinet again.
F09 (invented author): Keep both F08 beats as proposals only. I have not approved Ivo discovering the mark, or Mara choosing to tell him. The identity of the bell thief remains undecided. Neither Ivo giving the key nor the rejected Mara-steals-key branch proves who stole the bell.
F10 (invented author): Draft 8 is the current working version. In this fictional project, chapters 1 through 4 are available manuscript material; chapter 5 has only a provisional outline. These source turns do not include chapter prose, and no handover based only on these turns should claim to have inspected the manuscript files.
F11 (invented author): The next task is to inspect contradictions against these decisions and propose optional next beats. Do not draft chapter 5 or resolve the thief. If a claim depends on the chapter prose rather than these source turns, request that prose before assessing it. Keep new suggestions separate from canon until I approve them.
F12 (invented author): I have reviewed the decision summary for this fictional exercise: Ivo gives Mara the brass key in chapter 3; it opens the archive cabinet; the bell theft is Thursday; Mara sees the concealed mark in chapter 4 and Ivo has not discovered it by that chapter ending. Draft 8 is current, chapters 1 through 4 are the reported available scope, and chapter 5 is provisional. The mark meaning and thief identity remain undecided. F08 stays proposed. This confirms the decisions in these source turns, not any unseen chapter text or future beat.
Copy this, then replace the prompts with the details from your conversation.
The Bell at Greyharbor: worked handover Source boundary This is an original invented example, not a thredly summary or an actual conversation. All author and collaborator voices are fictional. Only F01-F12 are supplied: no chapter prose or manuscript files were read. Draft and chapter availability are reported in those excerpts, not verified file contents. Current canon - C01: Working title: The Bell at Greyharbor. [F01] - C03: Ivo willingly gives Mara the brass key in chapter 3. Mara does not steal it. [F04; confirmed F12] - C05: The bell was stolen on Thursday. No calendar date is specified. [F07; confirmed F12] - C07: The brass key opens the archive cabinet. Do not treat it as a key to the lighthouse door. [F05; confirmed F12] - C08: Mara sees a concealed mark on the inside of that cabinet in chapter 4. Seeing it does not establish its meaning. [F06; confirmed F12] Corrections and rejected branches - C02 / branch A: Mara stealing Ivo's brass key was an unapproved early brainstorm in F01. F04 rejects it and establishes C03. It was never approved canon. - C04 -> C05: Tuesday was author approved in F03, then explicitly superseded by Thursday in F07. Record one bell theft with a corrected day, not two thefts or two active timelines. - C06: The lighthouse-door use was a proposal in F02, rejected in F05. C07 is the approved archive-cabinet use. - C10: Ivo noticing the mark in chapter 4 was a proposal in F02, rejected in F06. Keep C09 as the current knowledge boundary. Story chronology - The chapter 3 key handoff precedes the chapter 4 mark sighting in the stated chapter sequence. [C03, C08; F04, F06] - The bell theft is Thursday, with no specified calendar date or chapter placement. Its timing relative to the handoff and mark sighting is not established. [C05; F07] - The chapter 5 discovery is a proposed future beat, not an established later event. [C11; F08-F09] Character knowledge at the end of chapter 4 - Mara: Has seen the concealed cabinet mark. The author has not established that she understands its meaning. [C08, C14; F06] - Ivo: Has not seen or discovered the mark by the chapter 4 ending. A chapter 5 proposal does not change this state. The sources do not establish other beliefs or suspicions he might have. [C09; F06, F12] Proposed future beats - P01 / C11: Ivo discovers the mark in chapter 5 when Mara shows him the cabinet. Proposed in F08; kept proposed in F09 and F12. Has not occurred in the established chapter 1-4 canon. - P02 / C12: Mara decides whether to tell Ivo before opening the cabinet again. Proposed in F08; unapproved in F09 and F12. Her decision has not been established. Version and manuscript scope - C15: Draft 8 is current. [F10, F12] - C16: Chapters 1-4 are reported available in the fictional project. No chapter text is included here, so this handover cannot certify scene-level consistency across those chapters. [F10, F12] - C17: Chapter 5 has only a provisional outline. The proposed beats above are not approved events or a completed chapter. [F08-F10, F12] Unresolved questions - C13: Who stole the bell? Undecided. Do not infer a thief from Ivo's key handoff or the rejected Mara theft branch. [F01, F09, F12] - C14: What does the concealed mark mean? Undecided. Do not infer understanding from Mara seeing it. [F06, F12] Next permitted task Inspect contradictions against the source decisions and propose optional next beats. Do not draft chapter 5 or resolve the thief. Ask for chapter prose before making any claim that depends on it; keep new suggestions proposed until the author approves them. [C18-C19; F11] Questions for your own story 1. Does each approved claim and chapter cutoff still match the current draft? 2. Has any proposal been approved since the quoted turns? If so, provide the approval source and add a dated change instead of silently rewriting the ledger. 3. What chapter excerpts should be supplied before the next contradiction pass?
Copy this, then replace the prompts with the details from your conversation.
Reader verification exercise Use only F01-F12. Classify each statement as author approved, rejected, superseded, proposed, unresolved, or unsupported. Give the source IDs and correct any stale claim before checking the key. 1. The bell was stolen on Tuesday. 2. The bell was stolen on Thursday. 3. Mara stole the brass key from Ivo in chapter 3. 4. The brass key opens the lighthouse door. 5. Ivo has discovered the concealed mark by the end of chapter 4. 6. Ivo discovers the concealed mark in chapter 5 when Mara shows him the cabinet. 7. Mara is the bell thief. 8. This handover was checked against the manuscript files for chapters 1-4. Verification key 1. Superseded: Tuesday was approved in F03 and replaced by Thursday in F07. Keep the correction history; do not add another theft. [C04 -> C05] 2. Author approved: Thursday replaces Tuesday in F07 and is confirmed in F12. No calendar date or placement relative to the chapter scenes is given. [C05] 3. Rejected: this unapproved F01 brainstorm is rejected in F04. Approved canon is Ivo willingly giving Mara the key in chapter 3. [C02 -> C03] 4. Rejected: F02 proposes lighthouse access; F05 rejects it and approves the archive cabinet. [C06 -> C07] 5. Rejected as a proposal and contradicted by approved canon: F02 proposes Ivo noticing it in chapter 4; F06 rejects that and states he has not seen or discovered it by that chapter ending. F12 confirms the boundary. [C10, C09] 6. Proposed: F08 suggests it; F09 keeps it unapproved and F12 retains F08 as proposed. Do not treat it as a completed event or present knowledge. [C11 / P01] 7. Unresolved thief identity; the assertion that Mara is the thief is unsupported. F09 and F12 explicitly leave the identity undecided; the rejected key-theft brainstorm supplies no evidence of the bell thief. [C13] 8. Unsupported inspection claim: F10 explicitly says the turns contain no chapter prose and forbids claiming manuscript-file inspection. F12 confirms only source-turn decisions. Reported availability is separate from inspected evidence. [C16, C19] Before using your own handover, repeat this check against your actual source excerpts and ask the author to review the corrections and unknowns.
Copy this, then replace the prompts with the details from your conversation.
Read the supplied source excerpts and this handover. Check each claim against its source ID and exact quote. Treat explicit author decisions and corrections as authority; do not assume that the latest suggestion is approved. List contradictions, unsupported claims, stale decisions, branch leakage, and character-knowledge errors with source IDs. If a check requires missing chapter prose, say what prose is needed. Keep any optional new beats labeled proposed, preserve unresolved questions, and stay within the stated next task. Ask the author to approve any proposed canon change before using it as established fact.
No. It is a manual text record you review and provide yourself. It makes no promise that a tool will remember, load, or enforce it.
No. The Bell at Greyharbor and all its author and collaborator excerpts are an original invented example. No actual conversation, customer story, or generated summary is shown.
Keep it labeled proposed until the author approves it. A possible chapter 5 discovery does not change what a character knows at the end of chapter 4.
No. Reported chapter availability is different from reading chapter prose. Ask for the relevant excerpts before making scene-level consistency claims.
Link the old claim to its explicit replacement and keep the old claim labeled superseded. In this example, Thursday replaces Tuesday for one bell theft; its calendar date and placement relative to the chapter scenes remain unknown.