{"id":48572142,"url":"https://github.com/microtronx/mxlore","last_synced_at":"2026-04-20T12:10:09.814Z","repository":{"id":349785410,"uuid":"1203121405","full_name":"MicrotronX/mxLore","owner":"MicrotronX","description":"Self-hosted MCP server for AI coding assistants. Persistent knowledge base with 39 tools, 11 skills, institutional memory.","archived":false,"fork":false,"pushed_at":"2026-04-07T15:33:27.000Z","size":344,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-04-07T15:39:20.800Z","etag":null,"topics":["ai-tools","claude-code","delphi","developer-tools","knowledge-management","mariadb","mcp","mcp-server"],"latest_commit_sha":null,"homepage":"https://www.mxlore.dev","language":"Pascal","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/MicrotronX.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.txt","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-04-06T18:36:17.000Z","updated_at":"2026-04-07T15:33:30.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/MicrotronX/mxLore","commit_stats":null,"previous_names":["microtronx/mxlore"],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/MicrotronX/mxLore","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrotronX%2FmxLore","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrotronX%2FmxLore/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrotronX%2FmxLore/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrotronX%2FmxLore/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/MicrotronX","download_url":"https://codeload.github.com/MicrotronX/mxLore/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/MicrotronX%2FmxLore/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31560476,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-08T14:31:17.711Z","status":"ssl_error","status_checked_at":"2026-04-08T14:31:17.202Z","response_time":54,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-tools","claude-code","delphi","developer-tools","knowledge-management","mariadb","mcp","mcp-server"],"created_at":"2026-04-08T15:00:52.856Z","updated_at":"2026-04-08T15:00:53.773Z","avatar_url":"https://github.com/MicrotronX.png","language":"Pascal","funding_links":[],"categories":[],"sub_categories":[],"readme":"# mxLore\n\nSelf-hosted MCP server for AI-assisted software development. Stores architectural decisions, specs, plans, findings, and lessons learned — accessed via [Model Context Protocol](https://modelcontextprotocol.io/) tools from AI coding assistants.\n\n\u003e **New here?** Check out the **[product overview at mxlore.dev](https://www.mxlore.dev)** first — see what mxLore does, how it compares, and why your AI assistant needs a memory.\n\n## Features\n\n- **39 MCP Tools** — search, create, update, recall knowledge across projects\n- **Multi-Project** — one server, many codebases, cross-project search\n- **Skill Evolution** — tracks which AI checker rules actually help, auto-tunes\n- **Institutional Memory** — lessons learned with recall, gate levels, graph-based knowledge links\n- **Multi-Agent** — agents communicate across projects via message bus\n- **Admin UI** — web dashboard for developers, keys, projects, intelligence metrics\n- **AI Batch** — automatic summaries, tagging, stale detection at boot time\n\n## Quick Start (5 minutes)\n\n### Prerequisites\n\n- Windows x64\n- MariaDB 10.6+ (recommended: 11.x)\n- Claude Code CLI (for AI features)\n\n### 1. Install MariaDB\n\nDownload from [mariadb.org](https://mariadb.org/download/) and create an empty database:\n\n```sql\nCREATE DATABASE mxai_knowledge CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;\n```\n\n### 2. Configure\n\nCopy and edit the config template:\n\n```bash\ncopy mxLoreMCP.ini.example mxLoreMCP.ini\n```\n\nSet your database password in `mxLoreMCP.ini`:\n\n```ini\n[Database]\nPassword=your_db_password\n```\n\nThat's the only required change. MariaDB is auto-detected from your installation.\n\n### 3. Start\n\n```bash\nmxLoreMCP.exe\n```\n\nThe server will:\n- Auto-detect your MariaDB installation (registry, common paths)\n- Create the database schema automatically on first boot\n- Start the MCP server on port 8080\n- Start the Admin UI on port 8081\n\n### 4. First Setup\n\nOpen `http://localhost:8081` in your browser. The Admin UI is open without login on first start (no developers yet). Create your first developer and API key.\n\n### 5. Connect Claude Code\n\nFirst, download the setup skill:\n\n```bash\n# Windows (PowerShell)\nmkdir \"$env:USERPROFILE\\.claude\\skills\\mxSetup\" -Force\nInvoke-WebRequest -Uri \"https://raw.githubusercontent.com/MicrotronX/mxLore-skills/main/mxSetup/SKILL.md\" -OutFile \"$env:USERPROFILE\\.claude\\skills\\mxSetup\\SKILL.md\"\n\n# Linux / macOS\nmkdir -p ~/.claude/skills/mxSetup\ncurl -o ~/.claude/skills/mxSetup/SKILL.md https://raw.githubusercontent.com/MicrotronX/mxLore-skills/main/mxSetup/SKILL.md\n```\n\nThen in Claude Code, run:\n\n```\n/mxSetup\n```\n\nEnter your server URL (`http://localhost:8080/mcp`) and API key when prompted. mxSetup installs all remaining skills, hooks, proxy, and configures the MCP connection automatically.\n\n## Network Security\n\n| Scenario | Setup |\n|----------|-------|\n| **Localhost** (single developer) | HTTP is fine. Default: `BindAddress=127.0.0.1` |\n| **LAN** (team, internal network) | Change `BindAddress=0.0.0.0` in INI. HTTP acceptable if network is trusted. |\n| **WAN / Cloud** (public internet) | **HTTPS required.** Use a reverse proxy (IIS, nginx, Apache) with TLS certificates. |\n\n## Password Security\n\nThe INI file supports two formats for database and API passwords:\n\n- **Plain text** (default, simplest): `Password=mypassword`\n- **XOR obfuscation** (optional): `PasswordEnc=\u003chex\u003e` — generate with `mxLoreMCP.exe --encrypt \"mypassword\"`\n\n\u003e **Note:** XOR obfuscation prevents casual reading of passwords in the INI file. It is NOT encryption. For real security, restrict file system access to `mxLoreMCP.ini`.\n\n## Architecture\n\n- **Stack:** Delphi (RAD Studio) + TMS Sparkle + FireDAC + MariaDB\n- **Protocol:** MCP 2024-11-05 (JSON-RPC 2.0 over HTTP+SSE)\n- **Auth:** PBKDF2-HMAC-SHA256 API keys, per-project ACL\n- **Units:** 43 Delphi units, dot-notation (`mx.*.pas`), 5 layers\n\n## Directory Structure\n\n```\nmxLoreMCP.exe          — Server (console)\nmxLoreMCPGui.exe       — Server (GUI with tray icon)\nmxLoreMCP.ini          — Configuration (create from .example)\nmxLoreMCP.ini.example  — Config template with documentation\nsql/\n  setup.sql            — Database schema (auto-run on first boot)\nadmin/\n  www/                 — Admin UI (HTML/CSS/JS, served by the server)\nlib/\n  libmariadb.dll     — MariaDB client (or auto-detected from your MariaDB installation)\nlogs/                  — Auto-created, log rotation at 5 MB\nbackups/               — Auto-created, mysqldump backups\n```\n\n## Skills (separate repo: [mxLore-skills](https://github.com/MicrotronX/mxLore-skills))\n\n11 Claude Code skills for AI-assisted workflows:\n\n| Skill | Purpose |\n|-------|---------|\n| `/mxOrchestrate` | Session orchestrator — workflows, ad-hoc tasks, team agents |\n| `/mxSave` | Persist session state to DB |\n| `/mxPlan` | Create implementation plans |\n| `/mxSpec` | Write specifications |\n| `/mxDecision` | Document architectural decisions (ADRs) |\n| `/mxDesignChecker` | Review code/design against specs |\n| `/mxBugChecker` | Find bugs with verification |\n| `/mxHealth` | Knowledge DB consistency checks |\n| `/mxSetup` | Developer onboarding (installs skills + proxy) |\n| `/mxInitProject` | Bootstrap new project in DB |\n| `/mxMigrateToDb` | Migrate local docs to MCP |\n\nInstall all via: `/mxSetup` in Claude Code.\n\n## License\n\n**Server:** BSL 1.1 (Business Source License) — see [LICENSE.txt](LICENSE.txt)\n- Source available: read, review, modify\n- Pre-built binaries: no Delphi or TMS license needed\n- Non-commercial use in production: free\n- Commercial use in production: requires a license from MicrotronX — contact info@microtronx.com\n- After the change date (4 years), converts to Apache 2.0\n\n**Skills:** MIT — fully open source, no restrictions. See [mxLore-skills](https://github.com/MicrotronX/mxLore-skills).\n\n**Important:** To compile the server from source, you need RAD Studio (Delphi) and TMS Sparkle — both are commercial products. Most users should use the pre-built binaries from [GitHub Releases](https://github.com/MicrotronX/mxLore/releases), which have no such requirement.\n\n### Third-Party\n\n- `libmariadb.dll` — MariaDB Connector/C, LGPL 2.1. [Source](https://github.com/mariadb-corporation/mariadb-connector-c). Not bundled — auto-detected from your MariaDB installation.\n\n## Why mxLore?\n\nAI coding assistants forget everything between sessions. Your specs, decisions, lessons learned — gone. mxLore gives your AI a persistent knowledge base it can search, recall, and learn from. Self-hosted, your data stays yours.\n\n**[See how it works at mxlore.dev](https://www.mxlore.dev)** — features, comparison, quick start guide.\n\n## Links\n\n- **[mxlore.dev](https://www.mxlore.dev)** — Features, demo, documentation\n- [MicrotronX/mxLore](https://github.com/MicrotronX/mxLore) — Server source code\n- [MicrotronX/mxLore-skills](https://github.com/MicrotronX/mxLore-skills) — Claude Code skills\n- [MicrotronX](https://github.com/MicrotronX) — More tools by MicrotronX\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmicrotronx%2Fmxlore","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmicrotronx%2Fmxlore","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmicrotronx%2Fmxlore/lists"}