README, docs and site: one story (the whole job, no avoidable rejections, the desktop app, your AI, install)

README is a short front page (about 120 lines) with one real visual per section; every former section moved
unchanged into docs/*.md with a docs/README.md index (journey, rejections, getting-started, commands, growth,
studios, store-art, listing-languages, api, skills, mcp, safety, faq, project). The site tells the same story on one
page: the teaser in the hero, real CLI output as text, the desktop app as a section with light/dark captures of eight
screens and the per-system downloads (desktop/latest.json); /desktop/ redirects to /#download. New capture tools:
tools/site-desktop-shots.ts (fixture-mode desktop frames) and tools/site-term-shots.ts (real CLI output as text and
WebP). tests/docs-links.test.ts checks README, docs and site links resolve.

cli/index.ts

@@ -23,7 +23,7 @@ ${Object.entries(COMMANDS).map(([k, [, d]]) => line([k, d])).join('\n')}
 ${COMMON}
 credentials: PLAY_SERVICE_ACCOUNT, ASC_KEY_ID, ASC_ISSUER_ID, ASC_P8, ANDROID_KEYSTORE(_PASSWORD) / ANDROID_KEY_ALIAS / ANDROID_KEY_PASSWORD, REVIEW_DEMO_USER / REVIEW_DEMO_PASSWORD (env or .storebound/credentials.json);
   profiles for many accounts: ~/.storebound/credentials/<profile>.json (same names), picked with --account <profile>
-docs: README.md and the skills/ folder of the toolchain`);
+docs: README.md, docs/ and the skills/ folder of the toolchain`);
 }
 
 /** Usage of one command: its header usage lines (or the table line), description and accepted options. */

cli/lib/agents.ts

@@ -1,6 +1,6 @@
 // Where each AI coding tool looks for Agent Skills and MCP servers in a project, and pure merge helpers that add the
 // storebound entry without touching anything else in those files. Formats checked against each tool's docs in
-// October 2026 (see README "Agent support").
+// October 2026 (see docs/project.md "Agent support").
 
 export type AgentId = 'claude' | 'codex' | 'cursor' | 'copilot' | 'windsurf' | 'gemini' | 'opencode';
 

cli/lib/declarations.ts

@@ -16,7 +16,7 @@ export const DATA_SAFETY_UPLOADED = '.storebound/data-safety.uploaded.csv';
 /**
  * Play Data safety "Account creation" (question PSL_SUPPORTED_ACCOUNT_CREATION_METHODS): the options Play Console
  * shows (Data safety > Data collection and security > Account creation) and the response ids storebound writes.
- * Google publishes no list of these ids (README "Not verified"): compare with a Console export (App content > Data
+ * Google publishes no list of these ids (docs/project.md "Not verified"): compare with a Console export (App content > Data
  * safety > Export to CSV) when an import is refused.
  */
 export const ACCOUNT_CREATION: Record<string, string> = {

cli/lib/desktop-dist.ts

@@ -11,7 +11,7 @@ import { homedir } from 'node:os';
 import { join } from 'node:path';
 
 export const BUNDLE_ID = 'dev.storebound.desktop';
-export const PAGE = 'https://storebound.dev/desktop/';
+export const PAGE = 'https://storebound.dev/#download';
 export const MANIFEST: Record<'stable' | 'beta', string> = {
   stable: 'https://storebound.dev/desktop/latest.json',
   beta: 'https://storebound.dev/desktop/beta/latest.json',

cli/shots/glyphs.ts

@@ -4,7 +4,7 @@
 // (uncoveredChars) and measures, in the page, which face drew every character (measureFonts). `shots check` turns a
 // glyph drawn by a fallback the config did not choose, or by a system font, into an ERROR.
 //
-// Default stack per script (the template defaults, documented in the README and the store-screenshots skill):
+// Default stack per script (the template defaults, documented in docs/store-art.md and the store-screenshots skill):
 //   Latin incl. Latin Extended-A / B and Vietnamese, Cyrillic, Greek: the configured title / body font
 //     (defaults: game Paytone One + Montserrat, app Inter + Inter; all four cover Latin Extended and Vietnamese;
 //     Inter and Montserrat cover Cyrillic too). A Cyrillic or Greek line switches to Noto Sans as a whole.