Stop Scrolling at 2 A.M.–Lyra Exporter Puts Every Claude & Gemini Chat in Your Pocket (Forever)

Because good prompts deserve better than an endless Cmd+F marathon.

01 The Mess—Why Your AI Chats Are Lost by Design

It’s 1:47 A.M. You know Claude sketched a micro-vs-serverless diagram last week, but the thread is buried under 300 newer talks. Gemini still holds half-finished React code you never copied out. Every platform is a silo; every search box is a black hole. Multi-AI productivity quickly turns into multi-tab paralysis.

02 The Fix—What Lyra Exporter Actually Does

1. Pull: a Tampermonkey script adds an EXPORT button inside Claude.ai, Gemini, NotebookLM and Google AI Studio.
2. Parse: a zero-backend React app (GitHub Pages-hosted or localhost) reads the JSON and renders a timeline + branch view—images, artifacts, “thinking” blocks included.
3. Push: tick any subset, hit Export, grab a ZIP of standard Markdown—ready for Obsidian, Notion, GitBook or Hugo.

No server, no token, no analytics—your data never leaves the browser tab unless you explicitly download it.

03 Hands-On—Three Copy-Paste Steps

Now:

• Open Claude.ai → purple “Export” appears → choose Full account.
• JSON is POSTed to Lyra (or saved locally). Drag-and-drop into the reader.
• Search: has:image, thinking:yes, artifact:react … hits highlight instantly.
• Star or mark “important”, then Export Operated → claude_2025-10-15_041234.zip (Markdown + /assets folder).

Benchmark: 480 threads, 73 images → 42 MB ZIP in 6.8 s on a 2022 MacBook Air.

04 Pro Tips—Branching, Filtering & CI Automation

• Branch visualisation is automatic: if you forked a prompt, Lyra draws a Git-like tree so you can follow every twist.
• Bulk filter: combine queries, e.g. star:true artifact:* to export only starred codepens.
• Nightly backup (road-mapped CLI v2.0) will let you cron:

lyra export --platform claude --output ./backup/$(date +%F).json
git add . && git commit -m "autobackup $(date +%F)" && git push

Until then, a 30-line Puppeteer script inside GitHub Actions does the same—see example repo in FAQ.

05 FAQ (Schema-Friendly HowTo)

Q1: My JSON is 150 MB—upload fails?
A: In-browser limit = 100 MB. Export by project first, or wait for CLI.

Q2: Will images break after weeks?
A: No. Export ZIP rewrites URLs to local /assets; keep the folder and links stay forever.

Q3: Can I import into SillyTavern or ChatGPT-Plus?
A: Lyra writes plain Markdown + front-matter; use markdown-to-json then feed any consumer API.

06 Take-Home Cheatsheet

• [ ] Tampermonkey + Lyra script installed and enabled
• [ ] Successfully exported ≥1 full-account JSON (<100 MB)
• [ ] Located at least one “image + thinking” thread via search
• [ ] Marked three chats “important” and batch-exported to Markdown
• [ ] Unzipped into Obsidian vault—images render without 404
• [ ] (Optional) Scheduled GitHub Action for daily headless backup

07 Quick Quiz

1. You star 93 out of 480 chats then export—what’s the typical size drop?
   Answer: 42 MB → 7.8 MB, an 81 % saving.

2. How do you turn an exported React artifact into an interactive demo?
   Answer: Feed the code to VitePress or Storybook, embed the build via <iframe>; let CI redeploy on every new export.

SEO Summary

Meta title (≤70)
Lyra Exporter: One-Click Backup & Search for Claude, Gemini, NotebookLM Chats

Meta description (≤160)
Open-source userscript + React app exports your entire Claude, Gemini, NotebookLM history to searchable Markdown in under 30 s—no server, no fees.

Keywords
Lyra Exporter, Claude export chat, Gemini backup JSON, AI conversation manager, export NotebookLM

Stop treating your best prompts like disposable lighters—back them up, search them, and build on them. Happy exporting!