jcode-works · jb-thery · Aug 24, 2026 · Aug 24, 2026
@@ -6,14 +6,15 @@
 [![Node.js](https://img.shields.io/node/v/@jcode.labs/ragmir)](https://www.npmjs.com/package/@jcode.labs/ragmir)
 [![AGPL-3.0](https://img.shields.io/github/license/jcode-works/jcode-ragmir)](./LICENSE)
 
-**Confidential local RAG for coding agents, scripts, and Node.js applications.**
+**Open-source confidential local RAG for coding agents, scripts, and Node.js applications.**
 
 Ragmir turns specifications, Word files, PDFs, spreadsheets, code, and local exports into cited
 evidence indexed and retrieved on your machine. Core works offline by default, never uploads your
 corpus, and calls no model. Ragmir is the local retrieval and evidence layer for agentic RAG
 workflows: the host agent plans and acts, while Ragmir supplies bounded, cited project evidence.
-Connect the agent or automation you already use through CLI, MCP, or a typed TypeScript API, or keep
-the complete workflow local with the optional Chat package.
+Choose the boundary that fits the project: keep the complete workflow local with a local consumer
+or the optional Chat package, or connect Claude Code, Codex, Kimi, OpenCode, Cline, or another
+compatible agent through CLI or MCP. In either mode, the corpus, index, and retrieval stay local.
 
 [Website](https://ragmir.com) · [npm](https://www.npmjs.com/package/@jcode.labs/ragmir) ·
 [Documentation](https://github.com/jcode-works/jcode-ragmir/wiki) ·

@@ -1,7 +1,7 @@
 {
   "$schema": "https://context7.com/schema/context7.json",
   "projectTitle": "Ragmir",
-  "description": "Confidential local retrieval and evidence for agentic RAG workflows, with offline cited retrieval. The TypeScript CLI, library, read-focused MCP server, and portable skills keep the corpus and index under ignored local state. Optional local Chat and TTS remain separate packages.",
+  "description": "Open-source confidential local retrieval and evidence for agentic RAG workflows, with offline cited retrieval. The TypeScript CLI, library, read-focused MCP server, and portable skills keep the corpus and index under ignored local state. Optional local Chat and TTS remain separate packages.",
   "excludeFolders": [
     "node_modules",
     "packages/ragmir-core/dist",

@@ -1,10 +1,10 @@
 # Agent integration
 
-Ragmir is the local retrieval and evidence layer for agentic RAG workflows. It indexes the selected
-project files locally and gives the AI or automation you choose cited passages through CLI or one
-stdio MCP server. It is not an autonomous agent, planner, or action runtime: planning, tool use, and
-action authority stay in the host. The default `local-hash` path keeps ingestion and retrieval
-offline. Core is model-agnostic, never uploads the corpus, and never calls a model itself.
+Ragmir is the open-source local retrieval and evidence layer for agentic RAG workflows. It indexes
+the selected project files locally and gives the AI or automation you choose cited passages through
+CLI or one stdio MCP server. It is not an autonomous agent, planner, or action runtime: planning,
+tool use, and action authority stay in the host. The default `local-hash` path keeps ingestion and
+retrieval offline. Core is model-agnostic, never uploads the corpus, and never calls a model itself.
 
 For an interactive repository-aware installation, paste the canonical prompt from the
 [quick-start guide](./quick-start.md) into the coding agent. It detects the package manager and
@@ -13,11 +13,15 @@ verifies the selected clients.
 
 Choose the handoff that matches the corpus:
 
-| Path | What stays local | What crosses the boundary |
+| Mode | What stays local | What crosses the boundary |
 | --- | --- | --- |
-| Preferred hosted AI | Corpus, index, and retrieval | Only returned passages, under the AI provider's data policy |
-| Local AI or automation | Corpus, index, retrieval, and the consumer | Nothing, unless that consumer uses another network service |
-| Ragmir Chat | Corpus, index, retrieval, and answer generation | One explicit model download during setup, then no network |
+| Preferred coding agents | Corpus, index, and retrieval | Only requested passages; a hosted provider's data policy applies |
+| Fully local | Corpus, index, retrieval, and the consumer | Nothing, unless the chosen consumer uses another network service |
+
+Fully local can mean a local CLI or MCP agent, deterministic automation without a model, or optional
+Ragmir Chat. Chat requires one explicit model download during setup, then generates on the
+workstation. Claude Code, Codex, Kimi, OpenCode, and Cline have generated native helpers; other
+compatible clients use the same CLI or stdio MCP contract.
 
 Prepare the target repository once:
 

@@ -1,7 +1,8 @@
 # Quick start
 
-Ragmir requires Node.js 22 or later. Choose the guided setup for a repository-aware installation,
-or use the manual commands below.
+Ragmir is open-source confidential local RAG and requires Node.js 22 or later. Choose the guided
+setup for a repository-aware installation or use the manual commands below. Core keeps the corpus,
+index, and retrieval local; connect your preferred coding agent or keep the consumer local too.
 
 <!-- ragmir-setup-prompt:start -->
 <details>

@@ -2,7 +2,7 @@
   "name": "jcode-ragmir",
   "version": "0.4.12",
   "private": true,
-  "description": "Monorepo for Ragmir, confidential local RAG with offline cited retrieval for coding agents and scripts, plus optional local Chat and TTS.",
+  "description": "Monorepo for Ragmir, open-source confidential local RAG with offline cited retrieval for coding agents and scripts, plus optional local Chat and TTS.",
   "type": "module",
   "license": "AGPL-3.0-only",
   "author": "Jean-Baptiste Thery (https://github.com/jb-thery)",

@@ -5,9 +5,9 @@
 [![Node.js](https://img.shields.io/node/v/@jcode.labs/ragmir)](https://www.npmjs.com/package/@jcode.labs/ragmir)
 [![AGPL-3.0](https://img.shields.io/npm/l/@jcode.labs/ragmir)](https://github.com/jcode-works/jcode-ragmir/blob/main/LICENSE)
 
-Confidential local RAG for coding agents and Node.js applications. Core indexes the project files
-you choose and retrieves bounded, cited evidence offline by default. It uploads no corpus, calls no
-LLM, and opens no HTTP port.
+Open-source confidential local RAG for coding agents and Node.js applications. Core indexes the
+project files you choose and retrieves bounded, cited evidence offline by default. It uploads no
+corpus, calls no LLM, and opens no HTTP port.
 
 [Project overview](https://github.com/jcode-works/jcode-ragmir#readme) ·
 [CLI](https://github.com/jcode-works/jcode-ragmir/blob/main/docs/cli-reference.md) ·

@@ -1,7 +1,7 @@
 {
   "name": "@jcode.labs/ragmir",
   "version": "0.4.12",
-  "description": "Confidential local RAG for coding agents, with offline cited retrieval through CLI, MCP, and TypeScript.",
+  "description": "Open-source confidential local RAG for coding agents, with offline cited retrieval through CLI, MCP, and TypeScript.",
   "type": "module",
   "license": "AGPL-3.0-only",
   "author": {

@@ -63,6 +63,8 @@ and omits the production sitemap.
 ## Public-copy contract
 
 - Lead with model-agnostic Core: cited local retrieval through a library, CLI, and MCP.
+- Place the two operating modes before the feature map: fully local, or bounded cited passages sent
+  to the user's preferred coding agents.
 - Present preferred cloud agents, local consumers, and model-free automation as clear choices.
 - Keep Chat, TTS, semantic embeddings, model downloads, Edge speech, OCR, and IndexNow explicit.
 - Explain team use positively and briefly: merge reviewed changes upstream, run `rgr team sync`,

@@ -1,24 +1,24 @@
 {
   "nav_library": "Install",
   "nav_features": "Features",
-  "nav_agents": "Agents",
+  "nav_agents": "Work modes",
   "nav_use_cases": "Use cases",
   "nav_team": "Team",
   "nav_menu": "Menu",
   "nav_menu_close": "Close menu",
   "nav_github": "GitHub",
   "nav_aria_label": "Main navigation",
   "language_label": "Language",
-  "seo_home_title": "Ragmir: local evidence layer for agentic RAG workflows",
+  "seo_home_title": "Ragmir: open-source local evidence for agentic RAG workflows",
   "seo_home_description": "Open-source TypeScript RAG library, CLI, and local MCP server: a private evidence layer for agentic RAG workflows. The host keeps action authority.",
   "seo_home_keywords": "Ragmir, confidential local RAG, TypeScript RAG library, coding agents, AI coding agents, cited retrieval, private specifications, monorepo documentation, PDF search, XLSX search, synced Drive files, local automation scripts, persistent Node.js RAG client, Model Context Protocol, MCP",
   "seo_author": "Jean-Baptiste Thery",
   "seo_robots": "index, follow",
   "seo_image_alt": "Ragmir confidential local RAG with cited retrieval for coding agents",
-  "hero_title": "Confidential local RAG for your coding agents.",
-  "hero_title_line_1": "Confidential local",
-  "hero_title_line_2": "RAG for your",
-  "hero_title_line_3": "coding agents.",
+  "hero_title": "Open-source confidential local RAG for your coding agents.",
+  "hero_title_line_1": "Open-source",
+  "hero_title_line_2": "confidential local RAG",
+  "hero_title_line_3": "for your coding agents.",
   "hero_description": "Ragmir turns specs, Word files, PDFs, and local exports into cited evidence for agentic RAG workflows. Retrieval stays local and bounded; the host keeps action authority.",
   "hero_primary_cta": "Install Ragmir",
   "hero_secondary_cta": "See use cases",
@@ -134,24 +134,20 @@
   "features_chat_item_2": "Choose verified Qwen or Gemma profiles.",
   "features_chat_item_3": "Use Metal, CUDA, or Vulkan where supported.",
   "features_chat_item_4": "Download during setup, then answer offline without a hosted API.",
-  "agents_eyebrow": "Agent workflows",
-  "agents_title": "The evidence layer for agentic RAG, not another memory system.",
-  "agents_text": "Connect Ragmir to the agent you prefer through CLI or MCP, or keep the complete workflow on the workstation. MCP starts with at most three compact citations by default; research may add three code matches. The agent expands only the evidence it needs. Ragmir retrieves evidence; the host keeps action authority. Core never depends on a specific model.",
+  "agents_eyebrow": "How to work with Ragmir",
+  "agents_title": "Keep everything local, or connect the coding agents you already use.",
+  "agents_text": "The corpus, index, and retrieval stay on your workstation in both modes. MCP starts with at most three compact citations by default; research may add three code matches. You choose whether those bounded passages remain local or go to a hosted agent. Ragmir retrieves evidence; the host keeps action authority.",
   "agents_command_label": "Agent setup command",
   "agents_command": "npx rgr install-agent --agents claude,codex,kimi,opencode",
   "agents_portable_title": "Move one frozen knowledge base",
   "agents_portable_text": "Export the active index, required model, embedded read-only runtime, skills, and MCP adapters. Verify it after moving; raw source files and logs stay out.",
   "agents_portable_command": "npx rgr portable export",
-  "agents_targets_title": "Choose where the evidence goes",
-  "agents_targets_text": "Send bounded cited passages to the agent you already use, keep the handoff local, use optional Chat for an offline answer, or consume retrieval results without any model.",
-  "agents_preferred_name": "Your preferred agent",
-  "agents_preferred_text": "Claude, Codex, Kimi, Hermes, and other compatible clients can consume cited passages through CLI or MCP. A hosted client's data policy applies to those passages.",
-  "agents_local_name": "Local agent",
-  "agents_local_text": "OpenCode or another local CLI or MCP consumer can keep retrieved passages on the workstation. No hosted Ragmir service is involved.",
-  "agents_offline_name": "Offline answers",
-  "agents_offline_text": "Optional Ragmir Chat generates cited answers from one verified local GGUF. Qwen and Gemma are included profile choices, not Core requirements.",
-  "agents_automation_name": "Apps and automation",
-  "agents_automation_text": "Local scripts, desktop tools, and CI jobs consume typed or JSON retrieval results with or without a generative model.",
+  "agents_targets_title": "Choose your operating mode",
+  "agents_targets_text": "Both modes start from the same open-source, model-agnostic Core and the same bounded, cited retrieval.",
+  "agents_full_local_name": "Fully local",
+  "agents_full_local_text": "Keep retrieval and consumption on the workstation with a local CLI or MCP client, deterministic automation, or optional Ragmir Chat using a verified GGUF. No hosted Ragmir service is involved.",
+  "agents_connected_name": "Your preferred coding agents",
+  "agents_connected_text": "Claude Code, Codex, Kimi, OpenCode, Cline, Hermes, and other compatible clients receive only the cited passages you request through CLI or MCP. A hosted client's data policy applies to those passages.",
   "use_cases_eyebrow": "Use cases",
   "use_cases_title": "Six local workflows built around realistic project files.",
   "use_cases_text": "From a portable OpenClaw base and private specifications to monorepos, synced Drive folders, media research, and administrative dossiers, each flow keeps indexing local, retrieves offline, and returns citations.",

@@ -1,24 +1,24 @@
 {
   "nav_library": "Installer",
   "nav_features": "Fonctions",
-  "nav_agents": "Agents",
+  "nav_agents": "Modes",
   "nav_use_cases": "Cas d'usage",
   "nav_team": "Équipe",
   "nav_menu": "Menu",
   "nav_menu_close": "Fermer le menu",
   "nav_github": "GitHub",
   "nav_aria_label": "Navigation principale",
   "language_label": "Langue",
-  "seo_home_title": "Ragmir : couche de preuves locale pour workflows RAG agentiques",
+  "seo_home_title": "Ragmir : preuves locales open source pour workflows RAG agentiques",
   "seo_home_description": "Bibliothèque TypeScript open source, CLI et serveur MCP local : couche de preuves privée pour les workflows RAG agentiques. L'hôte conserve l'autorité d'action.",
   "seo_home_keywords": "Ragmir, RAG local confidentiel, bibliothèque RAG TypeScript, agents de code, agents IA de code, recherche citée, spécifications privées, documentation monorepo, recherche PDF, recherche XLSX, fichiers Drive synchronisés, scripts d'automatisation locaux, client RAG Node.js persistant, Model Context Protocol, MCP",
   "seo_author": "Jean-Baptiste Thery",
   "seo_robots": "index, follow",
   "seo_image_alt": "Ragmir RAG local et confidentiel avec recherche citée pour agents de code",
-  "hero_title": "Un RAG local et confidentiel pour vos agents de code.",
-  "hero_title_line_1": "Un RAG local",
-  "hero_title_line_2": "et confidentiel pour",
-  "hero_title_line_3": "vos agents de code.",
+  "hero_title": "RAG open source, local et confidentiel pour vos agents de code.",
+  "hero_title_line_1": "RAG open source,",
+  "hero_title_line_2": "local et confidentiel",
+  "hero_title_line_3": "pour vos agents de code.",
   "hero_description": "Ragmir transforme les spécifications, fichiers Word, PDF et exports locaux en preuves citées pour les workflows RAG agentiques. La recherche reste locale et bornée ; l'hôte conserve l'autorité d'action.",
   "hero_primary_cta": "Installer Ragmir",
   "hero_secondary_cta": "Voir les cas d'usage",
@@ -134,24 +134,20 @@
   "features_chat_item_2": "Choisissez un profil Qwen ou Gemma vérifié.",
   "features_chat_item_3": "Utilisez Metal, CUDA ou Vulkan selon la machine.",
   "features_chat_item_4": "Téléchargez le modèle pendant la configuration, puis répondez hors ligne sans API hébergée.",
-  "agents_eyebrow": "Flux pour agents",
-  "agents_title": "La couche de preuves du RAG agentique, pas une mémoire de plus.",
-  "agents_text": "Branchez Ragmir sur l'agent de votre choix via CLI ou MCP, ou gardez tout le flux sur le poste. MCP commence avec au plus trois citations compactes par défaut ; la recherche peut ajouter trois extraits de code. L'agent développe uniquement les preuves utiles. Ragmir retrouve les preuves ; l'hôte conserve l'autorité d'action. Core ne dépend d'aucun modèle précis.",
+  "agents_eyebrow": "Comment travailler avec Ragmir",
+  "agents_title": "Gardez tout en local ou connectez les agents de code que vous utilisez déjà.",
+  "agents_text": "Le corpus, l'index et la recherche restent sur votre poste dans les deux modes. MCP commence avec au plus trois citations compactes par défaut ; la recherche peut ajouter trois extraits de code. Vous décidez si ces passages bornés restent locaux ou vont vers un agent hébergé. Ragmir retrouve les preuves ; l'hôte conserve l'autorité d'action.",
   "agents_command_label": "Commande d'intégration",
   "agents_command": "npx rgr install-agent --agents claude,codex,kimi,opencode",
   "agents_portable_title": "Déplacez une base documentaire figée",
   "agents_portable_text": "Exportez l'index actif, le modèle requis, le runtime embarqué en lecture seule, les skills et les adaptateurs MCP. Vérifiez le dossier après déplacement ; les sources brutes et journaux restent exclus.",
   "agents_portable_command": "npx rgr portable export",
-  "agents_targets_title": "Choisissez où vont les preuves",
-  "agents_targets_text": "Envoyez des passages cités et bornés à l'agent que vous utilisez déjà, gardez le transfert en local, utilisez Chat en option pour générer une réponse hors ligne ou consommez les résultats sans modèle.",
-  "agents_preferred_name": "Votre agent préféré",
-  "agents_preferred_text": "Claude, Codex, Kimi, Hermes et les autres clients compatibles consomment les passages cités via CLI ou MCP. La politique de données d'un client hébergé s'applique à ces passages.",
-  "agents_local_name": "Agent local",
-  "agents_local_text": "OpenCode ou un autre consommateur CLI ou MCP local garde les passages récupérés sur le poste. Aucun service Ragmir hébergé n'intervient.",
-  "agents_offline_name": "Réponses hors ligne",
-  "agents_offline_text": "Ragmir Chat, optionnel, génère des réponses citées avec un GGUF local vérifié. Qwen et Gemma sont des profils inclus, pas des exigences de Core.",
-  "agents_automation_name": "Applications et automatisations",
-  "agents_automation_text": "Les scripts locaux, les outils de bureau et les tâches CI consomment des résultats typés ou JSON, avec ou sans modèle génératif.",
+  "agents_targets_title": "Choisissez votre mode de fonctionnement",
+  "agents_targets_text": "Les deux modes partent du même Core open source et indépendant du modèle, avec la même recherche bornée et citée.",
+  "agents_full_local_name": "Entièrement local",
+  "agents_full_local_text": "Gardez la recherche et la consommation sur le poste avec un client CLI ou MCP local, une automatisation déterministe ou Ragmir Chat en option avec un GGUF vérifié. Aucun service Ragmir hébergé n'intervient.",
+  "agents_connected_name": "Vos agents de code préférés",
+  "agents_connected_text": "Claude Code, Codex, Kimi, OpenCode, Cline, Hermes et les autres clients compatibles reçoivent uniquement les passages cités demandés via CLI ou MCP. La politique de données d'un client hébergé s'applique à ces passages.",
   "use_cases_eyebrow": "Cas d'usage",
   "use_cases_title": "Six workflows locaux construits autour de fichiers projet réalistes.",
   "use_cases_text": "Qu'il s'agisse d'une base OpenClaw portable, de cahiers des charges privés, de monorepos, de dossiers Drive synchronisés, de recherche média ou de dossiers administratifs, chaque flux garde l'indexation en local, effectue la recherche hors ligne et renvoie des citations.",

@@ -2,7 +2,7 @@
   "name": "@jcode.labs/ragmir-landing",
   "version": "0.4.10",
   "private": true,
-  "description": "Static site for Ragmir, confidential local RAG with offline cited retrieval for coding agents and scripts.",
+  "description": "Static site for Ragmir, open-source confidential local RAG with offline cited retrieval for coding agents and scripts.",
   "type": "module",
   "license": "AGPL-3.0-only",
   "scripts": {