LISA Core - AI Memory Library

Semantic compression for AI conversations. Export from 9 AI platforms. Premium features incl. cloud sync and integrity verification.

🆕 v0.49.5 — Export Quality & Stability Cleaner exports: UI labels, timestamps, and tool-use noise stripped from all export formats (LISA-V, Raw JSON) No duplicate messages, fixed repeated content in Claude conversation exports Smarter semantic anchors: only meaningful signals captured, not every line of text Honest metrics, inference pre-resolution % now reflects real data, not estimates Ghost auto-save fixed: sporadic save prompts on tab close eliminated Daily limit naming, internal cleanup aligning code with actual behavior 🔹🔹🔹 Our AI Memory Library tool, Lisa Core, is all about semantic translation and memory library for AI conversations. Export from 9 platforms. One subscription unlocks LISA Core + Web App + (coming soon) Inference Shield. ✨AI Memory Library – LISA Core Your AI Conversations. Your Data. Your Control. What is LISA Core? LISA (Language Intelligence Semantic Anchoring) is a privacy-first browser extension that translates your AI conversations into machine-executable structured data. Just as a professional translator preserves meaning across languages, LISA preserves meaning across the human–AI boundary. The result: Any AI that reads a LISA file doesn't interpret — it executes. Every entity resolved. Every ambiguity eliminated. Every task tracked. Your conversations never leave your browser. Zero cloud processing. 100% local. ✨The Problem We Solve Your AI conversations are trapped — and misunderstood. ChatGPT won't let you continue in Claude Gemini doesn't know what you discussed in Grok Every platform locks your data in their silo When you paste raw conversation into a new AI, it parses, infers, and guesses — wasting compute on figuring out what you meant instead of acting on it You created that knowledge. You should own it — in a format machines can execute, not just read. ✨ The LISA Core Solution LISA Core as your Ai memory library extracts your conversations and translates them into portable, semantically structured files that work across ANY AI platform. Upload a LISA file to any AI platform once you saved from the supported AI platforms and instantly restore your full context with zero ambiguity. One conversation. Every platform. Forever yours. 🧠 Why Structured Semantic Data? Raw Text VS LISA Output AI must interpret meaning VS AI receives meaning directly Probabilistic parsing VS Deterministic execution Context can be lost VS Context is anchored and explicit "I think you meant…" VS "Executing SA001: security…" LISA translates your conversations from human language into machine language. Semantic Anchors (SA001…) — Topics the AI must address Action Vectors (AV001…) — Tasks the AI must track Reconstruction Protocol — How to rebuild context Meaning is fragile. Structure preserves it. 🌐 Supported Platforms (9 total) 🔜 Coming soon: Ollama local model support 🚀 Key Features 📤 One-Click Export Detects which AI platform you're using and extracts all messages with a single click. 🌍 Multilingual Support Auto-detects English, French, Spanish, German, Arabic, Chinese, Japanese, Korean, Russian, Hebrew — with RTL support for Arabic/Hebrew. 🔬 Semantic Enrichment LISA Core as you AI memory library doesn't just capture text — it translates structure into machine-executable format: Semantic Anchors (SA001…) with severity tags Action Vectors (AV001…) — extracted commands and tasks File detection and Git reference tracking Relationship mapping (API calls, imports, commit-to-file, branch-to-commit) 🧩 10-Category Ambiguity Resolution What LISA Core does: Scans your conversation and identifies 10 types of ambiguity, tagging them for resolution: Referential · Temporal · Priority · Completion · Social noise · Reasoning · Role · Dependency · State · Relevance What the LISA Web App adds: When you send a conversation through the Web App's semantic translation pipeline, ambiguity is resolved – pronouns become entities, "later" becomes timestamps, decisions are extracted as anchors. The output is machine‑ready, not just tagged. What Inference Shield will add (coming soon): Explicit, verifiable disambiguation with audit trails – for compliance, IP protection, and enterprise use. 🖥️ Claude Code Support Full extraction from terminal-based coding sessions — task summaries, code blocks, development context. ✂️ Selection Export Select specific text, right-click, and export just what you need. 🧠 Backend Semantic Translation (80:1 to 100:1) When processed through the LISA Web App, a 60,000-word conversation translates into ~700 semantic anchors. The file is smaller because ambiguity, redundancy, and noise are resolved — not because content is cut. 🧩 Three outputs in JSON and JSONL (if your AI model doesn't accept JSONL, simply rename the file extension to .json — it will upload and parse correctly) 🗜️ LISA JSON (Semantic) — Enriched export processed by LISA's semantic engine. Each message is analyzed for entities, concepts, relationships, and intent — pre-resolving ambiguity before 💾 Floating Save Button Two capture modes: 📄 Raw JSON — Full verbatim transcript, unprocessed. Every message as plain text. Use this for archiving, compliance records, or feeding into the LISA Web App for report generation. 📝LISA-V (Verbatim+) — Block-level JSONL that separates prose from code, assigns SHA-256 hashes to every code block, builds a Merkle root for tamper-proof integrity, and tracks open vs. resolved tasks. The most complete format for seamless session handoff to a fresh AI instance. 📚 Local Library All conversations organized locally. Browse, search, manage. Never lose an important conversation. 🔐 LISA Hash (Premium) SHA-256 cryptographic verification — perfect for legal documentation, research records, compliance. 🕒 Version History Git-style versioning. Compare changes, restore earlier versions, track evolution. 📖 How to Use 1. Capture — Floating LISA button or popup "Export Conversation" LISA Raw (JSON) — for archiving & reports LISA-V (JSONL) — for AI handoffs & context transfer Or right‑click selected text → "Export to LISA" 2. Save — Download or save to your local library 3. Continue Anywhere — Upload to any AI and say: "Read this and continue our conversation" Platform‑independent AI memory. 💡 Use Cases Developers — Save coding sessions, transfer context between AI assistants, document decisions. LISA-V preserves open vs. resolved tasks so the next AI instance knows exactly where you left off. Researchers — Cryptographic verification via Merkle root, audit trails, shareable records with tamper‑proof integrity. Professionals — Client documentation, compliance records, team context sharing. Generate Summary, Detailed, Executive, or Technical reports via the Web App. Translators & Multilingual Users — Auto-detection across 9 languages. Conversations in French, Arabic, German, Spanish, Chinese, Japanese, Korean, Russian are tagged and handled correctly. Students — Study sessions, learning continuity, personal knowledge library. Anyone Leaving ChatGPT — Take your conversation history with you. Your data is yours. 💰 Pricing — One Subscription, Full Suite Free Tier 5 exports per day 5 LISA-V saves per day 5 Raw JSON saves per day All 9 platforms supported Local library storage Semantic enrichment included Multilingual detection included Premium — $19/month (or $79/year — Save $149!) What you get — the complete LISA suite: 🔹 LISA Core Extension (Unlimited) Unlimited exports · Unlimited LISA-V/Raw JSON saves LISA Hash cryptographic verification Priority support · Advanced features 🔹 LISA Web App (Full Access) Link : https://lisa-web-backend-production.up.railway.app/landing Report generation (Summary, Detailed, Executive, Technical) with cryptographic audit trails Cloud library & share links — sync across devices, share with team Up to 12 semantic anchors — higher fidelity translation Extended context windows · Priority AI processing 🔹 Inference Shield (IP Protection) — Coming soon 3‑stage semantic obfuscation — sanitize sensitive data before sharing with AI Risk score analysis — identify potential IP leakage Safe-for-publication exports — share context without exposing proprietary information Adversarial AI pipeline — advanced protection for enterprise use All three products. One subscription. Complete AI governance. 🔒 Privacy First We can't see your data because we never receive it. ✅ 100% local processing in your browser ✅ Zero server uploads (unless you explicitly sync to Web App) ✅ No tracking, no analytics on your content ✅ Your conversations stay on YOUR device ✅ Open, portable JSON/JSONL format — no vendor lock‑in What we DO store (optional Web App sync): License keys (hashed with bcrypt) · Usage counters · Anchor certificates (SHA‑256 hashes, NOT your content) · Snapshot metadata (if you choose "Send to App") What we DO NOT store: ❌ Your original conversation text ❌ Your uploaded documents ❌ Your LISA JSON exports ❌ Any AI provider API responses In a world where AI companies harvest your conversations, LISA gives you control. Technical Specs Version: 0.49.5 Manifest: V3 Core Technology: Semantic translation with 10‑category ambiguity resolution Backend Translation: 80:1 to 100:1 semantic anchor generation Client-Side Enrichment: Entity extraction, concept weighting, relationship mapping, intent classification Hash: SHA‑256 (Premium) + Merkle root (LISA‑V) Storage: Chrome Local Storage Formats: JSON / JSONL Languages: EN, FR, ES, DE, AR, ZH, JA, KO, RU, HE (auto‑detected) Result: 95% of AI compute goes to reasoning, not 40% to resolving ambiguity. 🏢 About Built by SAT-CHAIN LLC. Founded by Amar Dahmani, a professional translator with 25 years of experience who understands that meaning is fragile. LISA applies the principles of professional translation to the human‑AI boundary: just as a skilled translator stabilizes meaning across languages, LISA stabilizes meaning across AI interfaces. Every other tool saves what was said. LISA saves what was meant. 📬 Contact: contact@sat-chain.com 🌐 Website: https://sat-chain.com ⚡ Get Started Install LISA Core and own your AI conversations. Your memory. Your data. Your control. © 2026 SAT-CHAIN LLC