tv shows to tv series

CONTEXT.md

@@ -0,0 +1,277 @@
+# TV Show Recommendation System (Self-Hosted, Multi-Agent)
+
+## 🎯 Purpose
+
+This document provides **complete context for AI agents and developers** to build and evolve a local, self-hosted TV show recommendation system.
+
+The system is designed to:
+
+* Prioritize **current user intent** over historical behavior
+* Use **multi-agent architecture**
+* Generate, evaluate, and organize recommendations
+* Handle **50–100 candidate series per request**
+
+---
+
+# 🧠 Core Principles
+
+1. **Context over history**
+
+   * Do NOT persist long-term taste profiles
+   * Every request must be evaluated independently
+
+2. **Generation ≠ Evaluation**
+
+   * Candidate generation and ranking must be separated
+
+3. **High recall → aggressive filtering**
+
+   * Generate many candidates (50–100)
+   * Filter and rank later
+
+4. **Structured communication between agents**
+
+   * All agents must exchange structured data (JSON-like)
+
+5. **Deterministic where possible**
+
+   * Interpretation and ranking should be consistent
+
+---
+
+# 🏗️ System Overview
+
+```
+User Input
+   ↓
+[1] Interpreter Agent
+   ↓
+[2] Retrieval Agent
+   ↓
+[3] Ranking Agent
+   ↓
+[4] Curator Agent
+```
+
+---
+
+# 1. 🧾 Interpreter Agent
+
+## Goal
+
+Convert raw user input into structured data.
+
+## Input
+
+Free-form user text describing:
+
+* Liked series
+* Disliked series
+* Preferences (themes, tone, characters)
+* Constraints (things to avoid)
+
+## Output Schema
+
+```json
+{
+  "liked": ["string"],
+  "disliked": ["string"],
+  "themes": ["string"],
+  "character_preferences": ["string"],
+  "tone": ["string"],
+  "avoid": ["string"]
+}
+```
+
+## Requirements
+
+* Normalize terminology (e.g., "spy" → "espionage")
+* Infer implicit preferences
+* Detect contradictions
+* Be deterministic (low temperature)
+
+---
+
+# 2. 🔎 Retrieval Agent
+
+## Goal
+
+Generate a **large and diverse pool (50–100)** of candidate TV series.
+
+## Strategy: LLM Generation
+
+* Generate candidate series from structured input
+* Focus on diversity and coverage (high LLM temperature)
+
+## Output Schema
+
+```json
+[
+  {
+    "title": "string",
+    "metadata": {
+      "themes": [],
+      "tone": [],
+      "tags": []
+    }
+  }
+]
+```
+
+## Constraints
+
+* Do NOT use external APIs
+* Favor recall over precision
+
+---
+
+# 3. ⚖️ Ranking Agent
+
+## Goal
+
+Categorize candidates into 4 confidence levels using relative comparison.
+
+## Categories
+
+* Definitely Like
+* Might Like
+* Questionable
+* Will Not Like
+
+## Method
+
+### Step 1: Pre-filter
+
+* Remove obvious mismatches
+* Enforce "avoid" constraints
+
+### Step 2: Pairwise Comparison
+
+* Compare candidates relative to each other and input
+
+### Step 3: Tagging
+
+* Assign each show to one of the four categories
+
+## Output Schema
+
+```json
+{
+  "definitely_like": ["title"],
+  "might_like": ["title"],
+  "questionable": ["title"],
+  "will_not_like": ["title"]
+}
+```
+
+## Requirements
+
+* Keep logic simple and consistent
+* Be deterministic (low temperature)
+
+---
+
+# 4. 🎯 Curator Agent
+
+## Goal
+
+Produce a clean, user-facing recommendation output.
+
+## Responsibilities
+
+* Group series by category
+* Provide short explanations
+* Ensure readability
+
+## Output Example
+
+```
+Definitely Like:
+- Show A → reason
+
+Might Like:
+- Show B → reason
+```
+
+---
+
+# 🔁 Feedback Loop (Re-Ranking Only)
+
+## Goal
+
+Improve ranking without storing long-term user preferences.
+
+## Inputs
+
+* External summaries, reviews, or insights
+* User-provided feedback from other sources
+
+## Behavior
+
+* Adjust ranking dynamically
+* Re-rank existing candidates
+* Do NOT persist user taste
+
+---
+
+# ⚙️ Agent Configuration
+
+## Temperature Guidelines
+
+* Interpreter: Low
+* Retrieval: Medium/High
+* Ranking: Low
+* Curator: Medium
+
+## Rules
+
+* Agents must be stateless
+* Pass all required context explicitly
+* No hidden memory
+
+---
+
+# 🧱 Project Structure
+
+```
+/agents
+  interpreter
+  retrieval
+  ranking
+  curator
+
+/pipelines
+  recommendation_flow
+```
+
+---
+
+# 🚫 Non-Goals
+
+* No persistent taste model
+* No long-term user profiling
+
+---
+
+# 🧭 Implementation Notes for AI Agents
+
+When extending or modifying this system:
+
+1. Do NOT introduce long-term memory of user preferences
+2. Do NOT merge agent responsibilities
+3. Always preserve structured input/output between agents
+4. Prefer simplicity over overengineering
+5. Ensure ranking remains interpretable and consistent
+
+---
+
+# ✅ Summary
+
+This system is a **stateless, multi-agent recommendation pipeline** focused on:
+
+* Strong alignment with current input
+* High candidate diversity
+* Structured filtering and ranking
+* Scalable handling of large candidate sets
+
+The architecture is intentionally simple, modular, and extensible.

README.md

@@ -1,6 +1,6 @@
 # Recommender
 
-A pure TypeScript monolith AI agent application that will recommend TV shows based on a very customized user profile and input.
+A pure TypeScript monolith AI agent application that will recommend TV series based on a very customized user profile and input.
 
 ## Project Structure
 

packages/backend/drizzle/0000_wild_joseph.sql

@@ -10,8 +10,8 @@ CREATE TABLE IF NOT EXISTS "recommendations" (
 	"id" uuid PRIMARY KEY DEFAULT gen_random_uuid() NOT NULL,
 	"title" text NOT NULL,
 	"main_prompt" text NOT NULL,
-	"liked_shows" text DEFAULT '' NOT NULL,
-	"disliked_shows" text DEFAULT '' NOT NULL,
+	"liked_series" text DEFAULT '' NOT NULL,
+	"disliked_series" text DEFAULT '' NOT NULL,
 	"themes" text DEFAULT '' NOT NULL,
 	"brainstorm_count" integer DEFAULT 100 NOT NULL,
 	"recommendations" jsonb,

packages/backend/drizzle/0003_changing_tv_series.sql

@@ -0,0 +1,2 @@
+ALTER TABLE "recommendations" RENAME COLUMN "liked_shows" TO "liked_series";
+ALTER TABLE "recommendations" RENAME COLUMN "disliked_shows" TO "disliked_series";

packages/backend/drizzle/meta/0000_snapshot.json

@@ -89,15 +89,15 @@
           "primaryKey": false,
           "notNull": true
         },
-        "liked_shows": {
-          "name": "liked_shows",
+        "liked_series": {
+          "name": "liked_series",
           "type": "text",
           "primaryKey": false,
           "notNull": true,
           "default": "''"
         },
-        "disliked_shows": {
-          "name": "disliked_shows",
+        "disliked_series": {
+          "name": "disliked_series",
           "type": "text",
           "primaryKey": false,
           "notNull": true,

packages/backend/src/agents/curator.ts

@@ -4,7 +4,7 @@ import { z } from 'zod';
 import { zodTextFormat } from 'openai/helpers/zod';
 
 const CuratorSchema = z.object({
-  shows: z.array(z.object({
+  series: z.array(z.object({
     title: z.string(),
     explanation: z.string(),
     category: z.enum(["Full Match", "Definitely Like", "Might Like", "Questionable", "Will Not Like"]),
@@ -24,7 +24,7 @@ export async function runCurator(
 ): Promise<CuratorOutput[]> {
   const mediaLabel = mediaType === 'movie' ? 'movie' : 'TV show';
 
-  const allShows = [
+  const allSeries = [
     ...(ranking.full_match ?? []).map((t) => ({ title: t, category: 'Full Match' as const })),
     ...ranking.definitely_like.map((t) => ({ title: t, category: 'Definitely Like' as const })),
     ...ranking.might_like.map((t) => ({ title: t, category: 'Might Like' as const })),
@@ -32,7 +32,7 @@ export async function runCurator(
     ...ranking.will_not_like.map((t) => ({ title: t, category: 'Will Not Like' as const })),
   ];
 
-  if (allShows.length === 0) return [];
+  if (allSeries.length === 0) return [];
 
   const canSearch = useWebSearch && supportsWebSearch;
   const preferenceSummary = `User preferences summary:
@@ -53,9 +53,9 @@ Rules:
 - cons: up to 3 short bullet points about what the user might not like based on their preferences
 - Be honest — explain why "Questionable" or "Will Not Like" ${mediaLabel}s got that rating`;
 
-  const chunks: typeof allShows[] = [];
-  for (let i = 0; i < allShows.length; i += CHUNK_SIZE) {
-    chunks.push(allShows.slice(i, i + CHUNK_SIZE));
+  const chunks: typeof allSeries[] = [];
+  for (let i = 0; i < allSeries.length; i += CHUNK_SIZE) {
+    chunks.push(allSeries.slice(i, i + CHUNK_SIZE));
   }
 
   const results: CuratorOutput[] = [];
@@ -66,14 +66,14 @@ Rules:
       temperature: 0.5,
       ...serviceOptions,
       ...(canSearch ? { tools: [{ type: 'web_search' as const }] } : {}),
-      text: { format: zodTextFormat(CuratorSchema, "shows") },
+      text: { format: zodTextFormat(CuratorSchema, "series") },
       instructions,
       input: `${preferenceSummary}
 
 ${mediaLabel}s to describe:
 ${showList}`,
     }));
-    results.push(...(response.output_parsed?.shows ?? []));
+    results.push(...(response.output_parsed?.series ?? []));
   }
 
   return results;

packages/backend/src/agents/interpreter.ts

@@ -15,8 +15,8 @@ const InterpreterSchema = z.object({
 
 interface InterpreterInput {
   main_prompt: string;
-  liked_shows: string;
-  disliked_shows: string;
+  liked_series: string;
+  disliked_series: string;
   themes: string;
   media_type: MediaType;
   feedback_context?: string;
@@ -43,8 +43,8 @@ Rules:
 - Be specific and concrete, not vague
 - For "requirements": capture explicit hard requirements the user stated that recommendations must satisfy — things like "must be from the 2000s onward", "must have subtitles", "must feature a female lead". Leave empty if no such constraints were stated.`,
     input: `Main prompt: ${input.main_prompt}
-Liked ${mediaLabel}s: ${input.liked_shows || '(none)'}
-Disliked ${mediaLabel}s: ${input.disliked_shows || '(none)'}
+Liked ${mediaLabel}s: ${input.liked_series || '(none)'}
+Disliked ${mediaLabel}s: ${input.disliked_series || '(none)'}
 Themes and requirements: ${input.themes || '(none)'}${feedbackSection}`,
   }));
 

packages/backend/src/agents/retrieval.ts

@@ -19,7 +19,7 @@ export async function runRetrieval(
   previousFullMatches: string[] = [],
 ): Promise<RetrievalOutput> {
   const mediaLabel = mediaType === 'movie' ? 'movie' : 'TV show';
-  const mediaLabelPlural = mediaType === 'movie' ? 'movies' : 'TV shows';
+  const mediaLabelPlural = mediaType === 'movie' ? 'movies' : 'TV series';
 
   const canSearch = useWebSearch && supportsWebSearch;
   const response = await parseWithRetry(() => openai.responses.parse({

packages/backend/src/agents/retrievalContinuous.ts

@@ -38,7 +38,7 @@ function buildSystemPrompt(
   seenTitles: string[]
 ): string {
   const mediaLabel = mediaType === 'movie' ? 'movie' : 'TV show';
-  const mediaLabelPlural = mediaType === 'movie' ? 'movies' : 'TV shows';
+  const mediaLabelPlural = mediaType === 'movie' ? 'movies' : 'TV series';
 
   let prompt = `You are a ${mediaLabel} recommendation specialist. Your task is to recommend titles that match the user's taste profile.
 

packages/backend/src/db/schema.ts

@@ -5,8 +5,8 @@ export const recommendations = pgTable('recommendations', {
   id: uuid('id').defaultRandom().primaryKey(),
   title: text('title').notNull(),
   main_prompt: text('main_prompt').notNull(),
-  liked_shows: text('liked_shows').notNull().default(''),
-  disliked_shows: text('disliked_shows').notNull().default(''),
+  liked_series: text('liked_series').notNull().default(''),
+  disliked_series: text('disliked_series').notNull().default(''),
   themes: text('themes').notNull().default(''),
   brainstorm_count: integer('brainstorm_count').notNull().default(100),
   media_type: text('media_type').notNull().default('tv_show'),

packages/backend/src/pipelines/continuous.ts

@@ -39,8 +39,8 @@ function mergeCuratorOutputs(a: CuratorOutput[], b: CuratorOutput[]): CuratorOut
 }
 
 interface ContinuousPipelineInput {
-  likedShows: string;
-  dislikedShows?: string;
+  likedSeries: string;
+  dislikedSeries?: string;
   themes?: string;
   requirements?: string;
   avoid?: string;
@@ -57,8 +57,8 @@ export async function runContinuousPipeline(
 ): Promise<CuratorOutput[]> {
   const startTime = Date.now();
   const {
-    likedShows,
-    dislikedShows = '',
+    likedSeries,
+    dislikedSeries = '',
     themes = '',
     requirements = '',
     avoid = '',
@@ -83,9 +83,9 @@ export async function runContinuousPipeline(
     const t0 = Date.now();
 
     interpreterOutput = await runInterpreter({
-      main_prompt: themes || 'recommend shows based on user preferences',
-      liked_shows: likedShows,
-      disliked_shows: dislikedShows,
+      main_prompt: themes || 'recommend series based on user preferences',
+      liked_series: likedSeries,
+      disliked_series: dislikedSeries,
       themes: themes,
       media_type: mediaType,
     });
@@ -267,8 +267,8 @@ export async function runContinuousPipeline(
       .set({
         title: aiTitle,
         main_prompt: themes || 'Continuous recommendations',
-        liked_shows: likedShows,
-        disliked_shows: dislikedShows,
+        liked_series: likedSeries,
+        disliked_series: dislikedSeries,
         themes: themes,
         brainstorm_count: totalCount,
         media_type: mediaType,

packages/backend/src/pipelines/recommendation.ts

@@ -201,8 +201,8 @@ export async function runPipeline(
     const t0 = Date.now();
     const interpreterOutput = await runInterpreter({
       main_prompt: rec.main_prompt,
-      liked_shows: rec.liked_shows,
-      disliked_shows: rec.disliked_shows,
+      liked_series: rec.liked_series,
+      disliked_series: rec.disliked_series,
       themes: rec.themes,
       media_type: mediaType,
       ...(feedbackContext !== undefined ? { feedback_context: feedbackContext } : {}),

packages/backend/src/routes/recommendations.ts

@@ -29,8 +29,8 @@ export default async function recommendationsRoute(fastify: FastifyInstance) {
   fastify.post('/recommendations', async (request, reply) => {
     const body = request.body as {
       main_prompt: string;
-      liked_shows?: string;
-      disliked_shows?: string;
+      liked_series?: string;
+      disliked_series?: string;
       themes?: string;
       brainstorm_count?: number;
       media_type?: string;
@@ -64,8 +64,8 @@ export default async function recommendationsRoute(fastify: FastifyInstance) {
       .values({
         title: title || 'Untitled',
         main_prompt: body.main_prompt ?? '',
-        liked_shows: body.liked_shows ?? '',
-        disliked_shows: body.disliked_shows ?? '',
+        liked_series: body.liked_series ?? '',
+        disliked_series: body.disliked_series ?? '',
         themes: body.themes ?? '',
         brainstorm_count,
         media_type,
@@ -85,8 +85,8 @@ export default async function recommendationsRoute(fastify: FastifyInstance) {
   // POST /recommendations/continuous — create record and run continuous pipeline with SSE
   fastify.post('/recommendations/continuous', async (request, reply) => {
     const body = request.body as {
-      liked_shows: string;
-      disliked_shows?: string;
+      liked_series: string;
+      disliked_series?: string;
       themes?: string;
       requirements?: string;
       avoid?: string;
@@ -108,8 +108,8 @@ export default async function recommendationsRoute(fastify: FastifyInstance) {
       .values({
         title,
         main_prompt: body.themes ?? 'Continuous recommendations',
-        liked_shows: body.liked_shows,
-        disliked_shows: body.disliked_shows ?? '',
+        liked_series: body.liked_series,
+        disliked_series: body.disliked_series ?? '',
         themes: body.themes ?? '',
         brainstorm_count: totalCount,
         media_type: mediaType,
@@ -155,8 +155,8 @@ export default async function recommendationsRoute(fastify: FastifyInstance) {
 
       // Run the continuous pipeline (it will now update the existing record)
       await runContinuousPipeline(recId, {
-        likedShows: body.liked_shows,
-        dislikedShows: body.disliked_shows ?? '',
+        likedSeries: body.liked_series,
+        dislikedSeries: body.disliked_series ?? '',
         themes: body.themes ?? '',
         requirements: body.requirements ?? '',
         avoid: body.avoid ?? '',
@@ -305,11 +305,11 @@ export default async function recommendationsRoute(fastify: FastifyInstance) {
       const feedbackContext =
         feedbackRows.length > 0
           ? feedbackRows
-              .map(
-                (f) =>
-                  `${mediaLabel}: "${f.item_name}" — Rating: ${f.stars}/3 stars${f.feedback ? ` — Comment: ${f.feedback}` : ''}`,
-              )
-              .join('\n')
+            .map(
+              (f) =>
+                `${mediaLabel}: "${f.item_name}" — Rating: ${f.stars}/3 stars${f.feedback ? ` — Comment: ${f.feedback}` : ''}`,
+            )
+            .join('\n')
           : undefined;
 
       await runPipeline(rec, (event) => {

packages/backend/src/types/agents.ts

@@ -83,8 +83,8 @@ export interface ContinuousSession {
 
 export interface ContinuousStartRequest {
   mediaType: 'tv_show' | 'movie';
-  likedShows: string;
-  dislikedShows?: string;
+  likedSeries: string;
+  dislikedSeries?: string;
   themes?: string;
   requirements?: string;
   avoid?: string;

packages/frontend/src/api/client.ts

@@ -18,8 +18,8 @@ async function request<T>(path: string, options?: RequestInit): Promise<T> {
 
 export function createRecommendation(body: {
   main_prompt: string;
-  liked_shows: string;
-  disliked_shows: string;
+  liked_series: string;
+  disliked_series: string;
   themes: string;
   brainstorm_count?: number;
   media_type: MediaType;
@@ -68,8 +68,8 @@ export function deleteRecommendation(id: string): Promise<{ ok: boolean }> {
 }
 
 export function createContinuousRecommendation(body: {
-  liked_shows: string;
-  disliked_shows?: string;
+  liked_series: string;
+  disliked_series?: string;
   themes?: string;
   requirements?: string;
   avoid?: string;

packages/frontend/src/components/NewRecommendationModal.tsx

@@ -8,8 +8,8 @@ interface NewRecommendationModalProps {
   onClose: () => void;
   onSubmit: (body: {
     main_prompt: string;
-    liked_shows: string;
-    disliked_shows: string;
+    liked_series: string;
+    disliked_series: string;
     themes: string;
     requirements?: string;
     avoid?: string;
@@ -33,19 +33,19 @@ const MEDIA_OPTIONS: Array<{
   label: string;
   description: string;
 }> = [
-  {
-    type: 'tv_show',
-    icon: '📺',
-    label: 'TV Shows',
-    description: 'Serialized stories, limited series, and long-form comfort watches.',
-  },
-  {
-    type: 'movie',
-    icon: '🎬',
-    label: 'Movies',
-    description: 'Feature films, prestige cinema, and one-night picks.',
-  },
-];
+    {
+      type: 'tv_show',
+      icon: '📺',
+      label: 'TV series',
+      description: 'Serialized stories, limited series, and long-form comfort watches.',
+    },
+    {
+      type: 'movie',
+      icon: '🎬',
+      label: 'Movies',
+      description: 'Feature films, prestige cinema, and one-night picks.',
+    },
+  ];
 
 const MODE_OPTIONS: Array<{
   mode: GenerationMode;
@@ -53,27 +53,27 @@ const MODE_OPTIONS: Array<{
   badge: string;
   description: string;
 }> = [
-  {
-    mode: 'brainstorm',
-    label: 'Brainstorm',
-    badge: 'Best for variety',
-    description: 'Explore a broad pool of options, then rank and curate the strongest fits.',
-  },
-  {
-    mode: 'continuous',
-    label: 'Continuous',
-    badge: 'Best for deep search',
-    description: 'Generate recommendations in chained batches for a steadier, longer-running hunt.',
-  },
-];
+    {
+      mode: 'brainstorm',
+      label: 'Brainstorm',
+      badge: 'Best for variety',
+      description: 'Explore a broad pool of options, then rank and curate the strongest fits.',
+    },
+    {
+      mode: 'continuous',
+      label: 'Continuous',
+      badge: 'Best for deep search',
+      description: 'Generate recommendations in chained batches for a steadier, longer-running hunt.',
+    },
+  ];
 
 export function NewRecommendationModal({ onClose, onSubmit }: NewRecommendationModalProps) {
   const [step, setStep] = useState<'type' | 'mode' | 'form'>('type');
   const [mediaType, setMediaType] = useState<MediaType>('tv_show');
   const [generationMode, setGenerationMode] = useState<GenerationMode>('brainstorm');
   const [mainPrompt, setMainPrompt] = useState('');
-  const [likedShows, setLikedShows] = useState('');
-  const [dislikedShows, setDislikedShows] = useState('');
+  const [likedSeries, setLikedSeries] = useState('');
+  const [dislikedSeries, setDislikedSeries] = useState('');
   const [themes, setThemes] = useState('');
   const [requirements, setRequirements] = useState('');
   const [avoid, setAvoid] = useState('');
@@ -100,7 +100,7 @@ export function NewRecommendationModal({ onClose, onSubmit }: NewRecommendationM
   }, [loading, onClose]);
 
   const mediaLabel = mediaType === 'movie' ? 'Movie' : 'TV Show';
-  const mediaPluralLabel = mediaType === 'movie' ? 'movies' : 'shows';
+  const mediaPluralLabel = mediaType === 'movie' ? 'movies' : 'series';
 
   const handleSelectType = (type: MediaType) => {
     setMediaType(type);
@@ -119,15 +119,15 @@ export function NewRecommendationModal({ onClose, onSubmit }: NewRecommendationM
   const handleSubmit = async (e: Event) => {
     e.preventDefault();
     if (generationMode === 'brainstorm' && !mainPrompt.trim()) return;
-    if (!likedShows.trim()) return;
+    if (!likedSeries.trim()) return;
 
     setLoading(true);
     try {
       if (generationMode === 'brainstorm') {
         await onSubmit({
           main_prompt: mainPrompt.trim(),
-          liked_shows: likedShows.trim(),
-          disliked_shows: dislikedShows.trim(),
+          liked_series: likedSeries.trim(),
+          disliked_series: dislikedSeries.trim(),
           themes: themes.trim(),
           brainstorm_count: brainstormCount,
           media_type: mediaType,
@@ -142,8 +142,8 @@ export function NewRecommendationModal({ onClose, onSubmit }: NewRecommendationM
       } else {
         await onSubmit({
           main_prompt: '',
-          liked_shows: likedShows.trim(),
-          disliked_shows: dislikedShows.trim(),
+          liked_series: likedSeries.trim(),
+          disliked_series: dislikedSeries.trim(),
           themes: themes.trim(),
           requirements: requirements.trim(),
           avoid: avoid.trim(),
@@ -258,7 +258,7 @@ export function NewRecommendationModal({ onClose, onSubmit }: NewRecommendationM
 
               <div class="modal-type-footer">
                 <div class="modal-selection-summary">
-                  <span class="summary-pill">{mediaType === 'movie' ? 'Movies' : 'TV Shows'}</span>
+                  <span class="summary-pill">{mediaType === 'movie' ? 'Movies' : 'TV series'}</span>
                   <span class="summary-pill">{selectedMode?.label}</span>
                 </div>
               </div>
@@ -281,7 +281,7 @@ export function NewRecommendationModal({ onClose, onSubmit }: NewRecommendationM
 
             <form class="modal-form" onSubmit={handleSubmit}>
               <div class="modal-summary-strip">
-                <span class="summary-pill">{mediaType === 'movie' ? 'Movies' : 'TV Shows'}</span>
+                <span class="summary-pill">{mediaType === 'movie' ? 'Movies' : 'TV series'}</span>
                 <span class="summary-pill summary-pill--accent">{selectedMode?.label}</span>
                 <span class="summary-caption">
                   {generationMode === 'brainstorm'
@@ -307,28 +307,28 @@ export function NewRecommendationModal({ onClose, onSubmit }: NewRecommendationM
 
               <div class="modal-form-grid">
                 <div class="form-group">
-                  <label for="liked-shows">{mediaLabel}s you liked</label>
+                  <label for="liked-series">{mediaLabel}s you liked</label>
                   <input
-                    id="liked-shows"
+                    id="liked-series"
                     type="text"
                     class="form-input"
                     placeholder={mediaType === 'movie' ? 'e.g. Inception, The Godfather' : 'e.g. Breaking Bad, The Wire'}
-                    value={likedShows}
-                    onInput={(e) => setLikedShows((e.target as HTMLInputElement).value)}
+                    value={likedSeries}
+                    onInput={(e) => setLikedSeries((e.target as HTMLInputElement).value)}
                     required
                   />
                   <span class="form-help">A few strong examples help the pipeline lock onto your taste.</span>
                 </div>
 
                 <div class="form-group">
-                  <label for="disliked-shows">{mediaLabel}s you disliked</label>
+                  <label for="disliked-series">{mediaLabel}s you disliked</label>
                   <input
-                    id="disliked-shows"
+                    id="disliked-series"
                     type="text"
                     class="form-input"
                     placeholder={mediaType === 'movie' ? 'e.g. Transformers' : 'e.g. Game of Thrones'}
-                    value={dislikedShows}
-                    onInput={(e) => setDislikedShows((e.target as HTMLInputElement).value)}
+                    value={dislikedSeries}
+                    onInput={(e) => setDislikedSeries((e.target as HTMLInputElement).value)}
                   />
                   <span class="form-help">Optional, but useful when you want to steer away from common misses.</span>
                 </div>

packages/frontend/src/hooks/useRecommendations.ts

@@ -14,8 +14,8 @@ type GenerationMode = 'brainstorm' | 'continuous';
 
 interface CreateBody {
   main_prompt: string;
-  liked_shows: string;
-  disliked_shows: string;
+  liked_series: string;
+  disliked_series: string;
   themes: string;
   requirements?: string;
   avoid?: string;
@@ -58,8 +58,8 @@ export function useRecommendations() {
 
       if (body.generation_mode === 'continuous') {
         const result = await createContinuousRecommendation({
-          liked_shows: body.liked_shows,
-          disliked_shows: body.disliked_shows,
+          liked_series: body.liked_series,
+          disliked_series: body.disliked_series,
           themes: body.themes,
           requirements: body.requirements ?? '',
           avoid: body.avoid ?? '',
@@ -73,7 +73,7 @@ export function useRecommendations() {
         const result = await createRecommendation(body);
         id = result.id;
       }
-      
+
       await refreshList();
       setSelectedId(id);
       return id;

packages/frontend/src/pages/Home.tsx

@@ -13,8 +13,8 @@ export function Home() {
 
   const handleCreateNew = async (body: {
     main_prompt: string;
-    liked_shows: string;
-    disliked_shows: string;
+    liked_series: string;
+    disliked_series: string;
     themes: string;
     requirements?: string;
     avoid?: string;

packages/frontend/src/pages/Recom.tsx

@@ -183,8 +183,8 @@ export function Recom({ id }: RecomProps) {
 
   const handleCreateNew = async (body: {
     main_prompt: string;
-    liked_shows: string;
-    disliked_shows: string;
+    liked_series: string;
+    disliked_series: string;
     themes: string;
     brainstorm_count?: number;
     media_type: import('../types/index.js').MediaType;
@@ -234,16 +234,16 @@ export function Recom({ id }: RecomProps) {
                         <span class="rec-info-value">{rec.main_prompt}</span>
                       </div>
                     )}
-                    {rec.liked_shows && (
+                    {rec.liked_series && (
                       <div class="rec-info-row">
                         <span class="rec-info-label">Liked</span>
-                        <span class="rec-info-value">{rec.liked_shows}</span>
+                        <span class="rec-info-value">{rec.liked_series}</span>
                       </div>
                     )}
-                    {rec.disliked_shows && (
+                    {rec.disliked_series && (
                       <div class="rec-info-row">
                         <span class="rec-info-label">Disliked</span>
-                        <span class="rec-info-value">{rec.disliked_shows}</span>
+                        <span class="rec-info-value">{rec.disliked_series}</span>
                       </div>
                     )}
                     {rec.themes && (
@@ -254,7 +254,7 @@ export function Recom({ id }: RecomProps) {
                     )}
                     <div class="rec-info-row">
                       <span class="rec-info-label">Media</span>
-                      <span class="rec-info-value">{rec.media_type === 'tv_show' ? 'TV Shows' : 'Movies'}</span>
+                      <span class="rec-info-value">{rec.media_type === 'tv_show' ? 'TV series' : 'Movies'}</span>
                     </div>
                     {rec.use_web_search && (
                       <div class="rec-info-row">

packages/frontend/src/types/index.ts

@@ -17,8 +17,8 @@ export interface Recommendation {
   id: string;
   title: string;
   main_prompt: string;
-  liked_shows: string;
-  disliked_shows: string;
+  liked_series: string;
+  disliked_series: string;
   themes: string;
   media_type: MediaType;
   use_web_search: boolean;