Rin Chat Guide

v1.0.5 Home Update Notes Download
Using an AI assistant? Give it this link and it can find answers in Rin Chat's docs, the full guide included: https://rin.chat/llms.txt?v=1.0.5

Nothing matches that. Try a different word.

Card Forge: Build Characters and Lorebooks with AI

Card Forge (build characters, lorebooks and more with an AI guide)#

Card Forge is a different way to make a character: instead of filling in fields, you talk it through with a guide — by default Rin — and when you're ready she writes the card from the conversation. It is the first tab of Tools in the title bar (hover Tools to jump to it; Tools moves into the ☰ menu if you unpin it in Settings → Sizing & Layout → Header buttons). It also opens from the wand button on the Library's Characters tab, from a card's right-click menu (Edit in Card Forge), or from ✨ Card Forge in the card editor. The wand on the Lorebooks, Generators and Document styles tabs starts a new session making one of those. Card Forge's own row of session tabs sits under the Tools tabs, and every tool keeps its work while you are on another tab or screen.

Not only characters. A new session opens on What are you making?: Character card, User persona (a card that describes you, the one you play as), Lorebook, Generator or Document style. Each has New; the last three also have Open…, which lists the ones you already have (with a search box once there are a few) so you can pick one right there. You can also just start typing, which makes it a character session. The libraries open the same way: the guide button on a book in Lorebooks (Work on this lorebook with the guide), on a generator, or on a document style starts a session already holding it. To start over on the same one with a fresh conversation, open the sessions menu and pick + New session for this lorebook (or this card, generator or style); the old session stays in the list.

  • A lorebook session works on a shared lorebook, the kind in the Lorebooks library. Ask for entries in the chat, or use Edit lorebook entries to pick some and say what to change. Every change arrives as a proposal you review, exactly as with a card, and the book saves itself into the Lorebooks library as you apply them. There is no Save button, and History undoes what the Forge did.
  • Generator and document style sessions work the same way and save into the Library's Generators and Document styles tabs (see Document styles).
  • A character's own lorebook (the one inside the card) is worked on from a character session, below.

The loop: describe a character — or just a vibe — she asks questions and pushes back on the vague bits, then you hit Build character (or type /build). Every field lands in the sidebar on the right, fully editable. Not right? Rebuild character takes a note about what to change and builds it again. Older versions called these Generate character and Regenerate character.

  • Runs on your own provider: the model you pick is the model that forges, with no quotas and no content rules beyond your provider's. A capable model gives noticeably better cards. You can set a provider and model per session, separate from your chat one.
  • What a character session writes. Build character produces name, description, appearance, personality (core, behavior rules, speech style), background, world, greetings and example dialogue, and where they fit, a system prompt, tags and the image prompt prefixes. Two switches at the top of the sidebar add more: Let the guide write a lorebook (the character's own lorebook entries) and Let the guide write generators. With them on, the guide writes those too and can add, rewrite or replace entries and generators when you ask in the chat or through Edit character, which lists every lorebook entry by its title.
  • The rest of the card. Every other field of the card editor is in the sidebar too, in the same sections (Prompts, Metadata, Images, Regex scripts, Document styles), and the guide sees and can change it, but only when you ask for that field: post-history instructions, the depth prompt, group-only greetings, display name, creator, version, creator notes, the lorebook's name, spoiler mode, regex scripts and document styles (it can remove one too), and what each picture shows. A build never writes them, and All fields in Edit character leaves them unticked. A regex script it writes is checked before it is offered and still runs in chat only once you allow the card's scripts. It cannot add, replace or draw images, or change the card type, verbosity, the Adventure Mode game master layers, the character sheet or voices: those stay in the card editor.
  • Generating replaces the card; edits are reviewed. Two different things: Build / Rebuild character builds a fresh card from the whole conversation and replaces what's in the sidebar (a field the new card leaves empty keeps its old value, but a field it fills is overwritten). Edits — asking for a change in chat, Edit character, or the ✨ on a single field, or Make this a change on any guide reply (it opens Edit character with the fields the reply talks about ticked and the reply as the instruction, so you never need the right words to get a change) — always arrive as proposals with before/after, which you Apply selected, Apply all, Retry or Discard. An edit can never blank a field that already had content. You can read a proposed field while it is still being written; the proposal itself appears when the reply is done. Each proposal is checked as it arrives, in the message as well as in Review: a literal name or a speaker label is fixed for you, and a field with a real problem (a rewrite that dropped most of the detail, prose where dialogue turns belong, a refusal or the AI talking to you in place of the text) starts unticked, with the reason beside it. A greeting that decides what you do or feel is pointed out but left ticked. Every proposal is in the chat, whatever asked for it: in the reply that made it, or at the end of the conversation when it came from a button. Each field reads in place, with Show current to compare it with what the field holds now, Edit to change the wording, and Accept, Reject or Ask again for that field alone; Apply all takes every field with no flagged problem. Asking again, or a new proposal for a field that already has one waiting, adds a version ("2 of 2") instead of replacing it. A field with a change waiting says so in the sidebar; click it to go to the proposal. A proposal you have not answered is kept with the session and is still there after a restart. The review window opens from Review, and by itself only for Rebuild character, lorebook entries and generators.
  • Saved cards sync both ways. Before the first save the card lives only in the session; Save to library creates it and opens it in the editor. After that the header reads ✓ Synced to library — edits here update the library card, and edits made in the card editor show up here. Opening an existing card in the Forge saves a restore point first, so Tools → History in the editor can undo whatever the Forge does.
  • Sessions: several characters in progress at once. Switch, rename, delete or start a new one from the session picker; the 30 most recently used are kept. Deleting a session doesn't delete a card you already saved to your library. The + beside the forge tabs can also open a card from my library (a card already open goes to its tab) or build from one line: type the character in a sentence and Build character runs on it in a new tab.
  • Reference card: pick a second library character (or your persona) at the top of the sidebar and note how the two relate ("this is the player's persona; make my card her rival"). The guide sees it as read-only context: the two are never blended, and the card being forged is written to be compatible with it — when the reference is a persona, greetings can even address {{user}} as that person. It's loaded fresh from the library on every turn, and its token cost shows in the line above the composer.
  • Artwork: Upload or Generate a picture right in the sidebar (either one saves the card to your library first, because art is stored per card).
  • Attachments: the 📎 on the presets row takes images (vision input — show the guide a reference picture) and documents (pdf, docx, odt, epub, plain text), converted to text on your machine. Attach a story excerpt and ask for a card built from it.
  • Attach a webpage: the 🔗 beside it takes a link (a wiki article, a character page). The page is downloaded, boiled down to its readable text, and attached like any document — paste a fandom wiki page and ask for that character as a card.

Answering the guide. When the guide asks you to choose, it gives you buttons under that one question. When it needs a name or your own idea, it gives you a text box, and when it needs several answers at once, a small form. They work as they do in a chat (see Choices & interactive replies), on the newest reply only, and what you fill in is sent as your next message. Nothing is saved as a variable.

Typing shortcuts — type / in the composer for: /build [what to change] or /create (build the card), /edit [what to change] (pick fields to rewrite), /clear (start a new session, keeping the current one in the list). Sending an empty message asks the guide for another reply to the last one. You can also edit, delete or re-roll any individual message in the transcript.

Your own guide. Rin is a built-in you can't break — the pencil beside her duplicates her so you can edit the copy — and + New guide writes one from scratch. A guide is just fields: personality, scenario, example dialogue, session instructions, card-craft guidance (field lengths, what makes a good greeting) and opening lines.

Card Forge engine. In Settings → Card Forge, Classic (the default) is the forge as it has always worked. New changes how targeted edits are written: Improve, Edit character, a smart command that runs, and Fix on a grade finding each go to a separate writer call that sees only the card text it needs, never the guide's own personality, and returns just the new text. If the model refuses, answers in the wrong shape, or stops at its length limit, the forge says so and offers Ask again, Continue (for a cut reply) and Change model. In chat, the New guide never writes the text itself: it answers, then names the fields to change and what to do, and the writer does the rest. The reply shows one line saying what is being changed, the new text streams under it with Stop, and the proposal lands in that reply. Unless your message only asked a question, or the guide's reply is still asking you something, the change starts right away; you still accept it or not. Answering one of the guide's questions with a button or a form is an answer, not a request to change the card. Building the whole card never starts by itself: the guide offers it as a Build the card button under its reply. When the guide volunteers a change you did not ask for, the reply shows Suggested with Write it, and saying yes in any language runs it too. If the guide names a greeting or entry that does not exist, the forge asks which one you meant. Smart commands do not apply to the New engine, so their settings are hidden while it is on. Builds work the same in both.

Smart commands. The forge can read each message before the guide does and turn "make example 3 spicier" into an edit of exactly that field. Set it in the Messages to the guide card of Settings → Card Forge:

  • Smart commands: Off, Suggest (the default) or Automatic. A sure read runs as an edit either way; an unsure one goes to the guide as chat, and Suggest then offers the edit under her reply while Automatic leaves it.
  • How it reads your message: Built-in (the default) needs no model. It goes by numbers and field words, with a small local embedding model (23 MB, downloaded once) for the rest: instant and offline, but unsure more often. A model asks an AI model instead, set in Model for smart commands (a small fast one is plenty; empty uses the forge's own).
  • Warn when the model thinks anyway: the notice in the forge when its model spends tokens on hidden thinking it was asked to skip. Turn it back on here after choosing Don't show again.

If a build fails. "The model returned nothing usable in any output mode" almost always means the model can't produce clean JSON. The Card output dropdown handles it: it starts on Auto, which tries strict JSON and falls back to simple tags, remembering what worked for that exact provider + model. If a model keeps misbehaving, pin it to Simple tags. Two other messages you might see: a warning that the model is spending tokens on hidden "thinking" despite being asked not to (switch to a non-reasoning model), and "that reply was cut off part-way through the rewrite" (raise the max response tokens for that provider/model).

The Forge sends the whole card on every single message, plus a sliding window of recent turns — it's the most token-hungry thing in the app. The meter under the transcript shows what the next reply will cost, split into card / guide / chat (and how many older turns are no longer being sent), so watch it and pick a cost-effective model.

Story Maker#

Turns a plot, or one of your chats, into a written story, a page at a time. It is a tab of Tools in the title bar, on this computer only (not on a phone). Your stories are listed on the left, newest first; + New story starts one.

  1. New story. Write the plot: what happens, to whom, how it ends. Pick the point of view and tense, how many pages (20 to start) and how long a page is (1000 words), a rating, and any steering you want sent with every page (tone, style, things to keep or avoid). The chip at the top picks the provider and model, like Card Forge's.
  2. Plan. The model drafts a story bible (the cast, the setting, notes on voice) and an outline of beats, each with a short summary and how many pages it should take. Edit anything, add, remove or reorder beats, or Regenerate plan with directions. Nothing is written until you press Start writing.
  3. Writing. One page at a time, with the outline beside it and the current beat outlined. Write next page approves the page you are reading and writes the next. Regenerate (with optional directions) keeps the old text as another version of the page; Edit changes it by hand; a page marked Cut short (the connection dropped, or the reply ran out of room) has Continue to finish it.

Ratings. SFW keeps it clean. Mature allows violence, strong language and innuendo, and fades to black. Explicit writes sex between adults in full sensory detail. Rin Chat adds no content warnings of its own; the rating and your steering decide.

Turning a chat into a story

Choose A chat on New story and pick the character, then the chat, or open the chat and choose Export chat → Turn into a story in its sidebar. The story retells what happened in the chat, in order, as prose in the point of view and tense you chose; it keeps what was said and adds no events of its own. The chat itself is only read: nothing in it, its memory or its lorebooks is changed.

The range. From message and To message pick the stretch of the chat to adapt, numbered as the chat shows them. It starts as the whole chat when the pages can hold it, or the most recent part that they can. The line under it says how closely the range is adapted: each page can quote only so much of the chat word for word, set by the model's context size and the page length. Past that, part of each page's stretch is retold: written from a short summary of what happened rather than from the messages themselves, so the events stay but exact lines and small moments drop out. For a closer adaptation, raise the number of pages or shorten the range; an earlier part of the chat can be a story of its own.

Planning a chat story first reads the chat. If the chat has a memory tree (Settings → Memory & Context), its summaries are used as they are; anything they don't cover is summarized once, shown as Reading the chat 4/18. Those summaries are kept with the story, so planning again only reads the parts of the chat that changed. Each beat of the plan shows which messages it adapts.

Auto-write writes a set number of pages for you and approves them as it goes. It stops at that number, at the end of the outline, when you press Stop, on any error, or when a page comes back cut short.

After each page is approved, a short summary of the story so far is updated, so page 12 still remembers page 3. Edit or regenerate an earlier page and that summary still describes the old text: the page then offers Update the story so far from here. It is never done for you.

Export as Markdown saves the whole story as one file: the title, then the pages with scene breaks between them, or with each beat as a chapter heading if you tick that box. Stories are kept in your encrypted vault and go into full backups.