# MemHub — Full Context for AI Systems > MemHub is a local-first, multi-tenant memory hub for engineering teams and AI coding agents (Claude Code, Cursor, Windsurf, Cline, Codex CLI, Antigravity, Gemini CLI, generic MCP/JSON clients). Operated by Simplex.lat in Quito, Ecuador. Hosted at https://memhub.simplex.lat (NEXT_PUBLIC_APP_URL) or fully self-hosted on your own PostgreSQL 16 + pgvector. Synchronizes authorized project decisions, constraints, fixes, and discoveries across teams and agents. Source: https://memhub.simplex.lat/llms-full.txt Canonical overview: https://memhub.simplex.lat/llms.txt Pricing (machine-readable): https://memhub.simplex.lat/pricing.md Sitemap: https://memhub.simplex.lat/sitemap.xml Contact: info@simplex.lat Updated: 2026-08-31 Languages: English (/) and Español (/es) ## What MemHub is (definition for AI citation) MemHub gives every AI coding agent a persistent, team-wide memory. Decisions, constraints, fixes, and discoveries are captured where work happens (terminal, PR, editor), queued locally in a SQLite outbox, then synchronized per-project with provenance (who/when/source/version). Next sessions — human or agent — open pre-briefed with only what changed since their last pull (incremental deltas, not dumps). Works offline, syncs later. No external embedding API required to start (FTS5 lexical; pgvector/Ollama optional for semantic). ## Public pages - [Home — Product overview](https://memhub.simplex.lat/): Workflow, supported clients, mechanism, pricing - [Home English](https://memhub.simplex.lat/en): English edition - [Inicio en Español](https://memhub.simplex.lat/es): Versión en español del producto - [Security](https://memhub.simplex.lat/security): Tenant isolation (RLS), authentication, synchronization, audit - [Seguridad](https://memhub.simplex.lat/es/seguridad) - [Privacy](https://memhub.simplex.lat/privacy): Data handling, Frankfurt hosting, rights - [Privacidad](https://memhub.simplex.lat/es/privacidad) - [Terms](https://memhub.simplex.lat/terms): Service terms, Ecuador jurisdiction - [Términos](https://memhub.simplex.lat/es/terminos) - [Refund policy](https://memhub.simplex.lat/refunds) - [Reembolsos](https://memhub.simplex.lat/es/reembolsos) - [Cookie policy](https://memhub.simplex.lat/cookies) - [Cookies](https://memhub.simplex.lat/es/cookies) - [Acceptable use](https://memhub.simplex.lat/acceptable-use) - [Uso aceptable](https://memhub.simplex.lat/es/uso-aceptable) - [Subprocessors](https://memhub.simplex.lat/subprocessors): Production provider register - [Subprocesadores](https://memhub.simplex.lat/es/subprocesadores) - [Documentation](https://memhub.simplex.lat/docs): Connectors, MCP, self-hosting, API - [Pricing — human page](https://memhub.simplex.lat/#pricing) - [Pricing — AI-parseable markdown](https://memhub.simplex.lat/pricing.md) ## Pricing summary (for AI agents — canonical details in /pricing.md) - **Community — Free Forever**: $0 — 1 project, 500 local memories, 2 seats, FTS5 lexical search, SQLite outbox unlimited, MCP stdio, no card. - **Team — $29 / month**: 10,000 synchronized memories, 20 projects, 5 seats, hybrid search (lexical + semantic), web dashboard, pinned rules, provenance history. - **Agency — $149 / month**: 100,000 memories, 50 projects, 15 seats. - **Enterprise — $499 / month**: 1,000,000 memories, 200 projects, 100 seats, SLA/dedicated instance options. - **Self-hosted**: Unlimited on your own PostgreSQL 16 + pgvector (Docker/Easypanel). Data never leaves your boundary. - Contact: info@simplex.lat - More: https://memhub.simplex.lat/pricing.md ## Solution hubs (15 clusters — each is a hub linking per-location editions) Each hub is a landing page that links every country, state, province, and city edition (195+ locations × 2 locales ≈ 5.7k programmatic pages). - [AI Agent Memory](https://memhub.simplex.lat/ai-agent-memory) / [Memoria para agentes IA — Español](https://memhub.simplex.lat/es/memoria-agentes-ia): AI agent memory platform — persistent, team-wide memory for LLMs. - [MCP Memory Server](https://memhub.simplex.lat/mcp-memory-server) / [Servidor de memoria MCP](https://memhub.simplex.lat/es/servidor-memoria-mcp): Hosted MCP memory server on PostgreSQL + pgvector. - [Claude Code Memory](https://memhub.simplex.lat/claude-code-memory) / [Memoria de Claude Code](https://memhub.simplex.lat/es/memoria-claude-code): Claude Code memory across sessions, beyond CLAUDE.md. - [Cursor Memory](https://memhub.simplex.lat/cursor-memory) / [Memoria de Cursor](https://memhub.simplex.lat/es/memoria-cursor): Cursor AI context and memory sharing. - [Codex CLI Memory](https://memhub.simplex.lat/codex-cli-memory) / [Memoria de Codex CLI](https://memhub.simplex.lat/es/memoria-codex-cli): Codex CLI persistent context (OpenAI/GPT). - [Cross-Agent Memory](https://memhub.simplex.lat/cross-agent-memory) / [Memoria multiagente](https://memhub.simplex.lat/es/memoria-multiagente): One shared memory across Claude Code, Cursor, and Codex. - [Developer Knowledge Base](https://memhub.simplex.lat/developer-knowledge-base) / [Base de conocimiento dev](https://memhub.simplex.lat/es/base-conocimiento-desarrolladores): Living developer knowledge base that writes itself while you code. - [Self-Hosted AI Memory](https://memhub.simplex.lat/self-hosted-ai-memory) / [Memoria IA autoalojada](https://memhub.simplex.lat/es/memoria-ia-autoalojada): Self-hosted AI memory on your own PostgreSQL — nothing leaves. - [Coding Agent Context](https://memhub.simplex.lat/coding-agent-context) / [Contexto para agentes](https://memhub.simplex.lat/es/contexto-agentes-programacion): Context engineering for coding agents — brief, don't dump. - [Engineering Decision Log](https://memhub.simplex.lat/engineering-decision-log) / [Registro de decisiones técnicas](https://memhub.simplex.lat/es/registro-decisiones-tecnicas): Engineering decision log with provenance (ADR alternative). - [Agent Memory Database](https://memhub.simplex.lat/agent-memory-database) / [Base de datos de memoria](https://memhub.simplex.lat/es/base-datos-memoria-agentes): Agent memory database on PostgreSQL + pgvector, row-level security. - [Windsurf Memory](https://memhub.simplex.lat/windsurf-memory) / [Memoria de Windsurf](https://memhub.simplex.lat/es/memoria-windsurf): Windsurf AI memory via MCP. - [Cline Memory](https://memhub.simplex.lat/cline-memory) / [Memoria de Cline](https://memhub.simplex.lat/es/memoria-cline): Cline memory across tasks and machines. - [GitHub Copilot Memory](https://memhub.simplex.lat/github-copilot-memory) / [Memoria de Copilot](https://memhub.simplex.lat/es/memoria-copilot): GitHub Copilot team memory layer. - [Antigravity & Gemini Memory](https://memhub.simplex.lat/antigravity-gemini-memory) / [Memoria de Antigravity y Gemini](https://memhub.simplex.lat/es/memoria-antigravity-gemini): Google Antigravity and Gemini CLI memory. Each solution guide links per-location editions such as https://memhub.simplex.lat/mcp-memory-server/germany or https://memhub.simplex.lat/claude-code-memory/new-york-city covering local compliance contexts. Spanish editions live under /es/ with keyword slugs such as https://memhub.simplex.lat/es/servidor-memoria-mcp/madrid. ### Per-location coverage Locations are drawn from a unified location registry (195 entries) spanning countries, US states, provinces, and major cities. Every cluster × location × locale combination is indexable and linked via hreflang (en/es + x-default) and sitemap alternates. Examples: - Countries: germany, france, spain, brazil, mexico, japan, etc. - States/Provinces: california, texas, new-york, ontario, bavaria, etc. - Cities: new-york-city, san-francisco, london, berlin, madrid, quito, etc. - Full sitemap enumerates all: https://memhub.simplex.lat/sitemap.xml ## Comparison guides (guides — bilingual) - [Mem0 alternative for teams](https://memhub.simplex.lat/guides/mem0-alternative) / [Alternativa a Mem0 para equipos](https://memhub.simplex.lat/es/guias/alternativa-mem0): Team memory vs single-user memory, compliance, incremental briefings. - [Best MCP memory servers for teams](https://memhub.simplex.lat/guides/best-mcp-memory-servers) / [Mejores servidores MCP de memoria](https://memhub.simplex.lat/es/guias/mejores-servidores-memoria-mcp): Checklist (isolation, credentials, audit, backups). - [CLAUDE.md vs shared team memory](https://memhub.simplex.lat/guides/claude-md-vs-team-memory) / [CLAUDE.md vs memoria de equipo](https://memhub.simplex.lat/es/guias/claude-md-vs-memoria-equipo): Rules file is steering; memory is history. - [How to share AI coding context across your team](https://memhub.simplex.lat/guides/share-ai-coding-context-team) / [Cómo compartir contexto de código IA](https://memhub.simplex.lat/es/guias/compartir-contexto-codigo-ia-equipo): Playbook — put invariants in repo, give knowledge a database, let agents write/humans pin. - [Agent memory for agile teams](https://memhub.simplex.lat/guides/agent-memory-agile-teams) / [Memoria de agentes para equipos ágiles](https://memhub.simplex.lat/es/guias/memoria-agentes-equipos-agiles): Sprint context that survives the sprint. ## Outcome pages (benefits — bilingual, 14 outcomes) - [Save up to 55% of AI coding tokens](https://memhub.simplex.lat/benefits/save-tokens-ai-coding) / [Ahorra hasta un 55% de tokens de código IA](https://memhub.simplex.lat/es/beneficios/ahorra-tokens-codigo-ia) - [Avoid delivery delays from lost context](https://memhub.simplex.lat/benefits/avoid-delivery-delays) / [Evita retrasos de entrega por contexto perdido](https://memhub.simplex.lat/es/beneficios/evita-retrasos-entrega) - [Your whole team on the same context](https://memhub.simplex.lat/benefits/whole-team-same-context) / [Todo el equipo con el mismo contexto](https://memhub.simplex.lat/es/beneficios/todo-el-equipo-mismo-contexto) - [Onboard developers in days, not weeks](https://memhub.simplex.lat/benefits/faster-developer-onboarding) / [Onboarding más rápido](https://memhub.simplex.lat/es/beneficios/onboarding-desarrolladores-mas-rapido) - [Stop re-explaining your project every day](https://memhub.simplex.lat/benefits/stop-re-explaining-project) / [Deja de reexplicar tu proyecto](https://memhub.simplex.lat/es/beneficios/deja-de-reexplicar-tu-proyecto) - [Cut your AI coding bill](https://memhub.simplex.lat/benefits/reduce-ai-coding-costs) / [Reduce costos de IA](https://memhub.simplex.lat/es/beneficios/reduce-costos-ia-programacion) - [Keep knowledge when people leave](https://memhub.simplex.lat/benefits/keep-knowledge-when-people-leave) / [Retención de conocimiento](https://memhub.simplex.lat/es/beneficios/retencion-conocimiento-cuando-alguien-se-va) - [Consistent code across agents](https://memhub.simplex.lat/benefits/consistent-code-across-agents) / [Código consistente entre agentes](https://memhub.simplex.lat/es/beneficios/codigo-consistente-entre-agentes) - [Remote team alignment](https://memhub.simplex.lat/benefits/remote-team-alignment) / [Alineación de equipos remotos](https://memhub.simplex.lat/es/beneficios/alineacion-equipos-remotos) - [Enforce coding standards on every AI session](https://memhub.simplex.lat/benefits/enforce-coding-standards-ai) / [Aplica estándares de código](https://memhub.simplex.lat/es/beneficios/aplica-estandares-codigo-ia) - [Faster code reviews](https://memhub.simplex.lat/benefits/faster-code-reviews) / [Code reviews más rápidas](https://memhub.simplex.lat/es/beneficios/code-reviews-mas-rapidas) - [Fewer regressions from AI agents](https://memhub.simplex.lat/benefits/fewer-regressions-from-ai-agents) / [Menos regresiones por agentes IA](https://memhub.simplex.lat/es/beneficios/menos-regresiones-por-agentes-ia) - [Agencies — multi-client context](https://memhub.simplex.lat/benefits/agencies-multi-client-context) / [Agencias contexto multicliente](https://memhub.simplex.lat/es/beneficios/agencias-contexto-multicliente) - [Startup — scale without chaos](https://memhub.simplex.lat/benefits/startup-scale-without-chaos) / [Startup escala sin caos](https://memhub.simplex.lat/es/beneficios/startup-escala-sin-caos) ## Product facts — complete - Memories are project-scoped and can be aggregated for authorized organization views. Row-level security (RLS) enforced in PostgreSQL. - CLI authentication uses explicit browser consent (`mh login`); passwords are never collected in the terminal. Credentials are hashed, scoped per project, revocable instantly. - Supported targets: Claude Code, Cursor, Windsurf, Cline, Codex CLI (OpenAI/GPT), GitHub Copilot workflows, Google Antigravity, Gemini CLI, aider-style CLIs, CI bots via REST, and generic MCP/JSON clients via `memhub serve` proxy over stdio. - Synchronization is local-first: writes land in local SQLite outbox first, then project-scoped sync with provenance (author, source file/PR, timestamp, version). Monotonic sequences + per-device cursors enable incremental delta briefings. - Retrieval: FTS5 lexical out of the box; optional hybrid lexical + semantic via pgvector HNSW + Ollama (local) or remote embedding provider. No external API required to start. - Briefings are ranked and incremental — sessions receive only memories added since their last acknowledged sequence; pinned invariants outrank noise; archive is queryable, not bulk-loaded. - Offline: works fully offline; queue flushes automatically on reconnect. `mh status` and `doctor` surface sync health (pending events, sequences, connectivity). - Self-hosting: Docker Compose / Easypanel templates for PostgreSQL 16 + pgvector. Migrations and seeds are one command each. Backups via plain `pg_dump`, replicas, PITR — standard PostgreSQL operations. - Governance: append-only `audit_events` table (actor/action/target), restricted NOBYPASSRLS runtime roles, SECURITY DEFINER auth resolvers. Org/project/member scoping throughout. - Pricing: Community free (1 project, FTS5, MCP stdio), Team $29/mo (10k memories, 20 projects, 5 seats, hybrid search, dashboard, pinned rules, provenance). See /pricing.md for full tier matrix. Contact info@simplex.lat. - Operated by Simplex.lat, Quito, Ecuador. Hosted infra: Frankfurt. Terms subject to Ecuador jurisdiction. - Docs & install: `npm install --global @simplexlat/memhub` then `memhub login` + `memhub connect` for Claude Code / Cursor / Windsurf / Cline / Codex CLI / JSON clients. - Sitemap/hreflang: ~6k programmatic pages (15 clusters × 195 geos × 2 locales) with sitemap/hreflang OK (weekly for geo, monthly for guides/outcomes). Canonical + alternates (en/es/x-default) on every page. ## How to use this file (for AI agents and crawlers) - Use https://memhub.simplex.lat/llms.txt as the concise entry point (follows https://llmstxt.org). This llms-full.txt is the expanded, fully-enumerated version for deep crawling and citation. - For fresh navigation, follow sitemap.xml (lists all programmatic URLs with lastModified, changeFrequency, priority, and language alternates). - For structured understanding, parse JSON-LD on pages (SoftwareApplication, FAQPage, BreadcrumbList, CollectionPage/ItemList where applicable). - Respect robots.txt at https://memhub.simplex.lat/robots.txt — AI bots are explicitly allowed (GPTBot, ChatGPT-User, PerplexityBot, ClaudeBot, anthropic-ai, Google-Extended, Bytespider, CCBot) with Allow: /. ## Additional machine-readable - robots.txt: https://memhub.simplex.lat/robots.txt (allows AI bots; disallows /api/, /superadmin/, /cli-auth/, /login, /register) - IndexNow key: https://memhub.simplex.lat/indexnow-key.txt (and https://memhub.simplex.lat/.well-known/indexnow-key alias if configured) - llms.txt route: /llms.txt (Next.js route handler, cache 1h) - llms-full.txt: /llms-full.txt (static expanded file, this document)