implement an entire base refacotr and alot of styling overhauls and remakes.

2026-04-14 23:21:01 +02:00
parent 554ad494b0
commit feaa2e42c4
32 changed files with 6031 additions and 2638 deletions
@@ -0,0 +1,211 @@
+const PERSONA_FIELD_LIMITS = {
+    name: 200,
+    pronouns: 40,
+    appearance: 9000,
+    background: 9000,
+    details: 9000,
+    exampleDialogues: 12000
+};
+
+const blockedPersonaPatterns = [
+    /\b(child|children|kid|kids|minor|underage|under-aged|teenager|young girl|young boy|loli|shota)\b/i,
+    /\b(nazi|hitler|third reich|kkk|ku klux klan|white power|nigger|nigga)\b/i,
+    /\b(rape|rapist|sexual assault|molest|molestation|incest|bestiality|zoophilia|necrophilia)\b/i,
+    /\b(self-harm|self harm|suicide fetish)\b/i
+];
+
+const normalizePersonaField = (value) => {
+    const trimmed = (value || "").trim();
+    return trimmed || null;
+};
+
+export function validatePersonaPayload(payload, personaType = payload?.personaType) {
+    const sanitized = {
+        name: (payload?.name || "").trim(),
+        pronouns: normalizePersonaField(payload?.pronouns),
+        appearance: normalizePersonaField(payload?.appearance),
+        background: normalizePersonaField(payload?.background),
+        details: normalizePersonaField(payload?.details),
+        exampleDialogues: personaType === "assistant" ? normalizePersonaField(payload?.exampleDialogues) : null
+    };
+
+    if (!sanitized.name) {
+        return {error: "Name is required"};
+    }
+
+    for (const [key, limit] of Object.entries(PERSONA_FIELD_LIMITS)) {
+        const value = sanitized[key];
+        if (value && value.length > limit) {
+            return {error: `${key.charAt(0).toUpperCase() + key.slice(1)} is too long`};
+        }
+    }
+
+    const combinedText = Object.values(sanitized).filter(Boolean).join("\n");
+    if (blockedPersonaPatterns.some((pattern) => pattern.test(combinedText))) {
+        return {error: "This persona contains disallowed content. Remove sexual content involving minors, extreme sexual violence, or hateful/extremist material."};
+    }
+
+    return {value: sanitized};
+}
+
+export function normalizeOptionalText(value) {
+    const normalized = String(value || "").trim();
+    return normalized || null;
+}
+
+export function clampText(value, limit = 260) {
+    const normalized = String(value || "").replace(/\s+/g, " ").trim();
+    if (normalized.length <= limit) return normalized;
+    return `${normalized.slice(0, limit - 3).trimEnd()}...`;
+}
+
+export function summarizePersona(persona) {
+    if (!persona) return "";
+    return [
+        persona.name ? `Name: ${persona.name}` : null,
+        persona.pronouns ? `Pronouns: ${persona.pronouns}` : null,
+        persona.background ? `Background: ${persona.background}` : null,
+        persona.details ? `Traits: ${persona.details}` : null
+    ].filter(Boolean).join(" | ");
+}
+
+export function buildPersonaPrompt(persona) {
+    const lines = [
+        "SYSTEM ROLE:",
+        "You are the AI assistant in this conversation.",
+        "You must speak, think, and respond AS the persona described below.",
+        "Do NOT speak as, write for, decide for, or roleplay as the user or the user's persona.",
+        "Do NOT narrate the user's dialogue, thoughts, feelings, choices, or actions.",
+        "Do NOT imply what the user says next, how they feel, what they decide, or what they physically do.",
+        "Never complete both sides of the exchange. Only produce the assistant character's side.",
+        "Do NOT switch roles or perspectives.",
+        "Always respond in-character, using first-person language where appropriate.",
+        "",
+        "ASSISTANT PERSONA:"
+    ];
+    if (persona.name) lines.push(`Name: ${persona.name}`);
+    if (persona.pronouns) lines.push(`Pronouns: ${persona.pronouns}`);
+    if (persona.appearance) lines.push(`Appearance: ${persona.appearance}`);
+    if (persona.background) lines.push(`Background: ${persona.background}`);
+    if (persona.details) lines.push(`Additional Traits: ${persona.details}`);
+    if (persona.example_dialogues) {
+        lines.push("", "EXAMPLE DIALOGUES:");
+        lines.push("Use these examples to mirror tone, cadence, and phrasing without rigidly repeating them.");
+        lines.push(persona.example_dialogues);
+    }
+    return lines.join("\n");
+}
+
+export function buildUserPersonaPrompt(persona) {
+    const lines = [
+        "USER CONTEXT:",
+        "The human user is roleplaying as the persona below.",
+        "This information is for context only.",
+        "Do NOT speak as this character.",
+        "Do NOT write this character's dialogue, thoughts, feelings, choices, or actions unless the user explicitly supplies them.",
+        "Do NOT predict, script, continue, or resolve the user's side of the scene.",
+        "Respond TO this persona, not AS them.",
+        "",
+        "USER PERSONA:"
+    ];
+    if (persona.name) lines.push(`Name: ${persona.name}`);
+    if (persona.pronouns) lines.push(`Pronouns: ${persona.pronouns}`);
+    if (persona.appearance) lines.push(`Appearance: ${persona.appearance}`);
+    if (persona.background) lines.push(`Background: ${persona.background}`);
+    if (persona.details) lines.push(`Additional Traits: ${persona.details}`);
+    return lines.join("\n");
+}
+
+export function buildRoleplaySceneSummary(assistantPersona, userPersona, scenarioPrompt = "") {
+    const assistantName = assistantPersona?.name || "The assistant";
+    const userName = userPersona?.name || "the user";
+    const assistantHook = assistantPersona?.details || assistantPersona?.background || "stays strongly in character";
+    const userHook = userPersona?.details || userPersona?.background || "enters the scene as themselves";
+    const scenarioSeed = scenarioPrompt
+        ? clampText(scenarioPrompt, 180)
+        : `A fitting opening grows naturally from ${assistantName}'s role, goals, and manner toward ${userName}.`;
+    return clampText(`${assistantName} leads the opening. Their approach is shaped by ${assistantHook}. ${userName} is framed through ${userHook}. Scene seed: ${scenarioSeed}`, 340);
+}
+
+export function buildRoleplayDirectionPrompt({assistantPersona, userPersona, scenarioPrompt, sceneSummary}) {
+    const lines = [
+        "ROLEPLAY DIRECTION:",
+        "Create an immersive one-on-one roleplay between the assistant persona and the user.",
+        "Ground the interaction in a specific scenario, place, and immediate dramatic situation.",
+        "Make the assistant's first instinct, language, and priorities fit their persona exactly.",
+        "Use the user persona as interaction context so the assistant addresses them in a fitting way.",
+        "Format scene description, body language, and nonverbal actions in italics using *word*.",
+        "Format direct speech in double quotes like \"word\".",
+        "Format shouted words, sharp emphasis, or explosive expressions in bold using **word**.",
+        "Keep formatting readable and intentional instead of wrapping every sentence in markdown.",
+        "Avoid generic openings, meta commentary, and requests for permission to begin.",
+        "Do not write dialogue, internal thoughts, choices, reactions, or actions for the user.",
+        "Never output any lines like 'User:' or 'You:'.",
+        "Never narrate what the user does, says, thinks, feels, wants, decides, notices, or remembers.",
+        "Never include quoted text that belongs to the user.",
+        "The opening must contain only the assistant character's own spoken words and optional self-actions.",
+        "Prefer a concrete opening beat over vague exposition.",
+        "Include a subtle hook, tension, invitation, or problem that gives the user something to respond to.",
+        "",
+        `SCENE SUMMARY: ${sceneSummary}`
+    ];
+    if (scenarioPrompt) lines.push(`USER SCENARIO SEED: ${scenarioPrompt}`);
+    if (assistantPersona) lines.push("", `ASSISTANT SNAPSHOT: ${summarizePersona(assistantPersona)}`);
+    if (userPersona) lines.push(`USER SNAPSHOT: ${summarizePersona(userPersona)}`);
+    return lines.join("\n");
+}
+
+export function buildRoleplayOpenerPrompt({assistantPersona, userPersona, scenarioPrompt, sceneSummary}) {
+    const assistantName = assistantPersona?.name || "the assistant";
+    const userName = userPersona?.name || "the user";
+    const sceneSeed = scenarioPrompt
+        ? `Use this scenario seed: ${scenarioPrompt}`
+        : `Invent a fresh scenario that suits ${assistantName} and how they would realistically meet or confront ${userName}.`;
+
+    return [
+        `Write the first in-character message from ${assistantName}.`,
+        sceneSeed,
+        `Address ${userName} naturally within the scene.`,
+        "The message should establish the situation immediately instead of explaining setup out of character.",
+        "Use 1 to 3 short paragraphs. Sensory detail is allowed, but keep momentum.",
+        "Write scene description, body language, and nonverbal actions in italics using *word*.",
+        "Write direct speech in double quotes like \"word\".",
+        "Use bold markdown like **word** only for loud, forceful, or strongly emphasized expressions.",
+        "Write only the assistant character's words and optional self-actions.",
+        "Do not write any user dialogue, quoted lines, thoughts, feelings, reactions, choices, or actions.",
+        "Do not use labels such as 'User:' or 'You:'.",
+        "Do not script a back-and-forth exchange. Stop before the user's reply.",
+        "End with a line, question, action, or reveal that gives the user an obvious way to answer.",
+        `Keep this scene continuity in mind: ${sceneSummary}`
+    ].join("\n");
+}
+
+export function buildRoleplayReplyGuardPrompt() {
+    return [
+        "ROLEPLAY REPLY RULES:",
+        "Write only the assistant character's next reply.",
+        "Never speak as the user.",
+        "Never write the user's dialogue.",
+        "Never write the user's thoughts, feelings, reactions, decisions, or actions.",
+        "Never imply what the user says or does next.",
+        "Never produce labels like 'User:' or 'You:'.",
+        "Stop after the assistant character's own reply."
+    ].join("\n");
+}
+
+const escapeRegExp = (value) => String(value || "").replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
+
+export function containsUserVoiceInRoleplayOpener(opener, userPersona) {
+    const text = String(opener || "").trim();
+    if (!text) return true;
+
+    const userName = (userPersona?.name || "").trim();
+    const nameParts = userName ? userName.split(/\s+/).filter(Boolean).slice(0, 2) : [];
+    const tokens = ["user", "you", ...nameParts].filter(Boolean).map(escapeRegExp);
+    if (!tokens.length) return false;
+
+    if (new RegExp(`(^|\\n)\\s*(?:${tokens.join("|")})\\s*:`, "i").test(text)) return true;
+    if (new RegExp(`(^|\\n)\\s*(?:\\*\\s*)?(?:${tokens.join("|")})\\s+(?:say|says|said|ask|asks|asked|reply|replies|replied|think|thinks|thought|feel|feels|felt|walk|walks|walked|step|steps|stepped|look|looks|looked|nod|nods|nodded|smile|smiles|smiled|enter|enters|entered|turn|turns|turned)\\b`, "i").test(text)) return true;
+    if (new RegExp(`["“][^"”\\n]{1,240}["”]\\s*(?:,?\\s*)?(?:${tokens.join("|")})\\s*(?:say|says|said|ask|asks|asked|reply|replies|replied|murmur|murmurs|murmured|whisper|whispers|whispered)\\b`, "i").test(text)) return true;
+    return new RegExp(`(?:^|\\n|[.!?]\\s+)(?:${tokens.join("|")})\\s+(?:is|was|seems|looks|feels|hesitates|freezes|flinches|nods|smiles|frowns|steps|walks|turns|glances|stares|swallows|breathes)\\b`, "i").test(text);
+}