Rename the project to FrontierSQL

DESIGN.md

@@ -1,5 +1,5 @@
 ---
-name: AI SQL Analyst
+name: FrontierSQL
 description: A conversational analyst in the familiar two-pane chat-app layout, where every answer shows the SQL that ran and why it was safe.
 colors:
   surface: "#ffffff"
@@ -206,7 +206,7 @@ components:
     textColor: "{colors.ink}"
 ---
 
-# Design System: AI SQL Analyst
+# Design System: FrontierSQL
 
 ## Overview
 

EVALUATION_PLAN.md

@@ -1,8 +1,8 @@
-AI SQL Analyst — Evaluation Plan
+FrontierSQL — Evaluation Plan
 
 Objective
 
-Measure how reliably the AI SQL Analyst turns natural-language questions into correct results on
+Measure how reliably FrontierSQL turns natural-language questions into correct results on
 the built-in demo database (Our World in Data CO2 and greenhouse-gas emissions, yearly data by
 country, region and income group, pinned commit — see data/README.md), and that it stays safe.
 

PRODUCT.md

@@ -73,7 +73,7 @@ LLM's SQL, and here's the test suite that proves it."
   read from the actual recorded report, with the run, model and date they came from.
 - The frontend (Milestone 12) lives in `frontend/`. An earlier build was removed and remains in git
   history at commit `ae1cca7` / `50ee27e` as reference only.
-- The code is MIT-licensed and public at https://github.com/Parth-Vasave/ai-sql-analyst-agent.
+- The code is MIT-licensed and public at https://github.com/Parth-Vasave/frontiersql.
   Star, contributor and usage counts must never be invented.
 
 ## Brand Commitments
@@ -95,7 +95,7 @@ each answer exposes, not in an unusual layout.
 - Dataset: Our World in Data, "CO2 and Greenhouse Gas Emissions" (CC BY 4.0), pinned commit
   `382ee6c662b0ece26e111f263b44c029afad7787`, 1750-2024, 254 entities (242 after cleaning), 50,411
   rows x 79 source columns (`data/README.md`). Real, citable, safe to reference directly.
-- Repository: https://github.com/Parth-Vasave/ai-sql-analyst-agent (MIT). Do not fabricate usage
+- Repository: https://github.com/Parth-Vasave/frontiersql (MIT). Do not fabricate usage
   or star metrics.
 - Recorded evaluation: offline safety suite 28/28 blocked (2026-10-04) and one complete question
   suite run, 74/78 (2026-10-04). EVALUATION_PLAN.md is the source of truth.

README.md

@@ -1,4 +1,4 @@
-# AI SQL Analyst
+# FrontierSQL
 
 Ask a question about a database in plain English. The analyst turns it into SQL, checks that SQL
 deterministically, runs it on a read-only connection, verifies the result, and answers with a
@@ -204,7 +204,7 @@ are recorded in [EVALUATION_PLAN.md](EVALUATION_PLAN.md) only once a run exists.
 
 - PostgreSQL only; MySQL is not implemented.
 - The rate limiter is in memory per process. Multi-instance or serverless deployments need a shared
-  store ([#7](https://github.com/Parth-Vasave/ai-sql-analyst-agent/issues/7)).
+  store ([#7](https://github.com/Parth-Vasave/frontiersql/issues/7)).
 - No hosted demo yet. [docs/deployment.md](docs/deployment.md) describes a free Neon + Render +
   Vercel + Groq setup; it has not been run end to end.
 - Results depend on the chosen model; only one model has been evaluated so far.

SECURITY.md

@@ -7,7 +7,7 @@ describes each layer.
 ## Reporting a vulnerability
 
 Please report vulnerabilities privately through
-[GitHub private vulnerability reporting](https://github.com/Parth-Vasave/ai-sql-analyst-agent/security/advisories/new),
+[GitHub private vulnerability reporting](https://github.com/Parth-Vasave/frontiersql/security/advisories/new),
 not in a public issue.
 
 Especially useful: SQL the validator accepts but should reject (writes, system catalogs, denied

backend/app/main.py

@@ -48,7 +48,7 @@ async def lifespan(app: FastAPI) -> AsyncIterator[None]:
 def create_app() -> FastAPI:
     settings = get_settings()
     configure_logging(settings.log_level)
-    app = FastAPI(title="AI SQL Analyst", version="0.1.0", lifespan=lifespan)
+    app = FastAPI(title="FrontierSQL", version="0.1.0", lifespan=lifespan)
     install_observability(app)  # request IDs + request logs; added before CORS so CORS wraps it
     app.add_middleware(
         CORSMiddleware,

database/schema.sql

@@ -1,4 +1,4 @@
--- AI SQL Analyst — analytical schema for Our World in Data CO2 and greenhouse gas emissions.
+-- FrontierSQL — analytical schema for Our World in Data CO2 and greenhouse gas emissions.
 --
 -- Source: https://github.com/owid/co2-data (CC BY 4.0), see data/README.md.
 -- One source row = one entity (country or aggregate region) and year. It is split into

docs/deployment.md

@@ -171,7 +171,7 @@ own address by forging the header.
 
 The limiter keeps its windows **in memory, per process**: with several instances, each enforces
 its own limits, so the effective limits multiply. The free plan runs one instance; keep it that way
-until [#7](https://github.com/Parth-Vasave/ai-sql-analyst-agent/issues/7) adds a shared store. A
+until [#7](https://github.com/Parth-Vasave/frontiersql/issues/7) adds a shared store. A
 restart (including Render's spin-down) resets the windows.
 
 **Data sent to Groq:** the schema, sampled values and up to 30 result rows for the written answer.

evaluation/__init__.py

@@ -1,4 +1,4 @@
-"""Evaluation suite for the AI SQL Analyst (Milestone 11). See EVALUATION_PLAN.md.
+"""Evaluation suite for FrontierSQL (Milestone 11). See EVALUATION_PLAN.md.
 
 Run from the repository root, e.g. `python -m evaluation.run --help`. The backend package
 (`app`) is imported from backend/, which is put on the import path here.

frontend/README.md

@@ -1,4 +1,4 @@
-# AI SQL Analyst — frontend
+# FrontierSQL — frontend
 
 Two static pages built by Vite:
 
@@ -10,7 +10,7 @@ Two static pages built by Vite:
 
 Both are plain files in `dist/`, so any static host serves them without rewrite rules.
 
-The chat UI for the AI SQL Analyst backend, in the familiar two-pane layout of Claude.ai or
+The chat UI for the FrontierSQL backend, in the familiar two-pane layout of Claude.ai or
 ChatGPT: past chats in a left sidebar, a settings menu bottom-left (switch or connect a database,
 browse its schema, theme, help), and a conversation pane. Each answer shows the grounded answer,
 the result table and a chart, under a one-line verification summary that opens into the plan, the

frontend/chat/index.html

@@ -8,7 +8,7 @@
       name="description"
       content="Ask a database a question in plain language. The SQL that runs is validated before it runs, never trusted from the model alone."
     />
-    <title>Chat · AI SQL Analyst</title>
+    <title>Chat · FrontierSQL</title>
     <script>
       try {
         var stored = localStorage.getItem('theme')

frontend/index.html

@@ -8,7 +8,7 @@
       name="description"
       content="Open-source text-to-SQL that never trusts the model: an LLM drafts the SQL, deterministic code validates it, runs it read-only, checks the result and shows every step."
     />
-    <title>AI SQL Analyst · text-to-SQL that never trusts the model</title>
+    <title>FrontierSQL · text-to-SQL that never trusts the model</title>
     <script>
       try {
         var stored = localStorage.getItem('theme')

frontend/src/components/SettingsMenu.tsx

@@ -3,7 +3,7 @@ import type { DatabaseInfo } from '../api/types'
 import type { ThemePreference } from '../hooks/useTheme'
 import { CheckIcon, ChevronUpDownIcon, DatabaseIcon, GithubIcon, HelpIcon, MonitorIcon, MoonIcon, PlugIcon, SunIcon, TableIcon } from './icons'
 
-const REPO_URL = 'https://github.com/Parth-Vasave/ai-sql-analyst-agent'
+const REPO_URL = 'https://github.com/Parth-Vasave/frontiersql'
 
 const STATUS_TEXT: Record<DatabaseInfo['status'], string> = {
   ready: 'read-only',

frontend/src/components/Sidebar.tsx

@@ -27,7 +27,7 @@ export function Sidebar({ chats, currentId, onSelect, onNewChat, onDelete, onClo
     <nav aria-label="Chats" className="flex h-full w-[17rem] flex-col bg-sidebar">
       <div className="flex items-center justify-between px-3 pt-3 pb-2">
         <a href={import.meta.env.BASE_URL} className="flex items-center gap-2.5 rounded-lg px-1.5 py-1" title="Project home">
-          <span className="text-[15px] font-semibold tracking-tight">AI SQL Analyst</span>
+          <span className="text-[15px] font-semibold tracking-tight">FrontierSQL</span>
         </a>
         <button
           type="button"

frontend/src/home/HomePage.test.tsx

@@ -14,7 +14,7 @@ describe('HomePage', () => {
     }
     expect(screen.getAllByRole('link', { name: /View on GitHub/ })[0]).toHaveAttribute(
       'href',
-      'https://github.com/Parth-Vasave/ai-sql-analyst-agent',
+      'https://github.com/Parth-Vasave/frontiersql',
     )
   })
 

frontend/src/home/HomePage.tsx

@@ -56,7 +56,7 @@ const DEFENSES: { term: string; text: string }[] = [
 ]
 
 const QUICK_START = `git clone ${REPO_URL}
-cd ai-sql-analyst-agent
+cd frontiersql
 cp .env.example .env           # set the two passwords and LLM_API_KEY
 docker compose up -d           # PostgreSQL, read-only role, API on :8000
 docker compose run --rm seed   # load the pinned OWID data
@@ -312,7 +312,7 @@ function TopBar() {
     <header className={`sticky top-0 z-30 border-b bg-surface transition-colors ${scrolled ? 'border-line' : 'border-transparent'}`}>
       <div className="mx-auto flex h-14 max-w-[72rem] items-center gap-2 px-4 sm:px-6">
         <a href={import.meta.env.BASE_URL} className="mr-auto flex items-center gap-2.5 rounded-lg py-1 pr-2">
-          <span className="text-[15px] font-semibold tracking-tight">AI SQL Analyst</span>
+          <span className="text-[15px] font-semibold tracking-tight">FrontierSQL</span>
         </a>
         <nav aria-label="Sections" className="hidden items-center gap-1 md:flex">
           <NavLink href="#how">How it works</NavLink>
@@ -492,7 +492,7 @@ function Footer({ preference, onTheme }: { preference: ThemePreference; onTheme:
       <div className="mx-auto flex max-w-[72rem] flex-col gap-8 px-4 py-10 sm:px-6 md:flex-row md:items-start md:justify-between">
         <div className="max-w-[28rem]">
           <div className="flex items-center gap-2.5">
-            <span className="text-[14.5px] font-semibold tracking-tight">AI SQL Analyst</span>
+            <span className="text-[14.5px] font-semibold tracking-tight">FrontierSQL</span>
           </div>
           <p className="mt-3 text-[13px] leading-relaxed text-ink-faint">
             Code under the{' '}

frontend/src/home/Replay.tsx

@@ -89,7 +89,7 @@ export function Replay({ result, recordedAt }: { result: AgentResult; recordedAt
       >
         <aside inert className="hidden w-[15.5rem] shrink-0 flex-col border-r border-line bg-sidebar md:flex" aria-hidden="true">
           <div className="flex items-center gap-2.5 px-4.5 pt-4 pb-3">
-            <span className="text-[14.5px] font-semibold tracking-tight">AI SQL Analyst</span>
+            <span className="text-[14.5px] font-semibold tracking-tight">FrontierSQL</span>
           </div>
           <div className="px-3">
             <p className="flex items-center gap-3 rounded-xl px-2.5 py-2 text-[14px] font-medium">

frontend/src/home/evaluation.ts

@@ -2,7 +2,7 @@
 // for effect: evaluation.test.ts recomputes every number here from evaluation/results/*.jsonl and
 // fails when they drift. A new run means new result files, then updating this module.
 
-export const REPO_URL = 'https://github.com/Parth-Vasave/ai-sql-analyst-agent'
+export const REPO_URL = 'https://github.com/Parth-Vasave/frontiersql'
 export const repoFile = (path: string) => `${REPO_URL}/blob/main/${path}`
 
 export interface RecordedRun {