✨ Cindy Glam
94%
@ to mention a file
Cindy can make mistakes. She makes a lot of mistakes about important stuff. But never about lip gloss. — tinyfiles.tv / Cindy
Template · Proposed · Not connected

⬇ Backend — what a backend could look like

Plausible plan for Cindy. We’ll fix the details later — this is the shape of it, not the final wiring. The chats + glam % you see now are local-only toys.

At a glance

💬 1 — Chats + messages

Real chats on disk / DB, not the chats[] array in this page. List, open, new, search — all via API.

✨ 2 — Cindy Glam

That 94% number becomes real: glam_events + compliments. +1% per cute chat is a toy — real one scores glam, not count.

💅 3 — Cindy engine

One server endpoint that calls the LLM, streams tokens, and keeps the “totally iconic” vapid personality. Browser never holds the API key.

💖 4 — Compliments

“Tell me I’m cute” becomes POST /api/compliment. Compliments are logged, so Cindy remembers what you liked — not auto-trusted for everything else.

👤 5 — Auth + users

Users, free plan, usage limits. So your Cindy isn’t everyone’s Cindy, and your chats stay yours.

🛡️ 6 — Safety + moderation

Filter mean corrections, rate-limit, log abuse. Cindy learns from you, not from anyone being mean about your selfie.

What it’ll take, step by step

  • Step 1 — Server: Deno / Node server. Serves this page + /api/*. WebSocket / SSE for streaming replies.
  • Step 2 — Chats: GET /api/chats lists, GET /api/chat/:id reads, POST /api/chat creates. Messages append, never overwrite.
  • Step 3 — Messages: POST /api/message saves your message, calls Cindy, streams her reply back, saves that too.
  • Step 4 — Glam: GET /api/glam returns %. POST /api/glam/event logs what Cindy obsessed over. No more glam = min(99, glam+1) in the browser.
  • Step 5 — Compliments: POST /api/compliment {chat_id, message_id, compliment} queues it. Approved = glam data.
  • Step 6 — Models: GET /api/models returns Cindy v1.0 / v1.1 / Pro. Switching models switches prompt + caps.

Effort — plausible, will fix later

PieceFilesPlausible effort
Server + chats APIserver.ts, routes/chats.ts~2–3 hrs
Messages + streamingroutes/messages.ts~2–3 hrs
Glam + complimentsroutes/glam.ts, routes/compliments.ts~2–3 hrs
Auth + users + capsauth.ts, db/*~3 hrs
Wire toy → realcindy-interface.html only~1 hr
Total~10 files~2 weekends

Tables — proposed

CREATE TABLE users (id TEXT PRIMARY KEY, email TEXT, plan TEXT, created_at DATETIME);
CREATE TABLE chats (id TEXT PRIMARY KEY, owner_id TEXT, title TEXT, model TEXT, created_at DATETIME);
CREATE TABLE messages (id INTEGER PRIMARY KEY, chat_id TEXT, role TEXT, content TEXT, created_at DATETIME);
CREATE TABLE compliments (id INTEGER PRIMARY KEY, chat_id TEXT, message_id INTEGER, compliment TEXT, status TEXT, created_at DATETIME);
CREATE TABLE glam_events (id INTEGER PRIMARY KEY, owner_id TEXT, kind TEXT, delta REAL, created_at DATETIME);
CREATE TABLE models (id TEXT PRIMARY KEY, name TEXT, prompt TEXT, cap_per_day INT);

API — proposed

GET  /api/chats                    -> [{id, title, model}]
GET  /api/chat/:id                 -> {chat, messages[]}
POST /api/chat {title, model}      -> 201 {id}
POST /api/message {chat_id, text}  -> stream (SSE) Cindy reply
POST /api/compliment {chat_id, message_id, compliment} -> 202 queued
GET  /api/glam                    -> {pct, events[]}
GET  /api/models                   -> [{id, name}]
🛠️ Fix later — open questions: Which LLM powers Cindy? How vapid is too vapid? Who reviews compliments? What stops a mean correction from making Cindy mean? What’s the per-user cost cap? None decided — modal is the placeholder until we pick.

Template only — chats + glam above are still local-only until this is wired.