By ojowwalker77
Build a persistent semantic memory for Claude across your codebase, enabling multi-source research with web/GitHub/docs, automated plugin diagnostics and fixes, dead code/circular dep detection, comprehensive reviews with security checks, manual reindexing, expert TypeScript/Swift guidance, memory stats/warnings, custom bash hooks, and local MCP servers for chat/docs.
npx claudepluginhub ojowwalker77/claude-matrix --plugin matrixThis skill should be used when the user asks to "research a topic", "deep research", "investigate thoroughly", "multi-source research", "comprehensive analysis", or needs to conduct research across web, documentation, and memory sources.
This skill should be used when the user asks to "run matrix diagnostics", "check matrix health", "fix matrix issues", "troubleshoot matrix", or mentions matrix not working. Runs comprehensive diagnostics on the Matrix plugin and automatically fixes issues.
This skill should be used when the user is working with Effect-TS, asks to "write Effect code", "use Effect", "functional TypeScript", "handle errors with Effect", "dependency injection Effect", "Effect Layer", or needs expert-level guidance on Effect-TS patterns, error handling, concurrency, and best practices.
This skill should be used when the user asks to "list matrix solutions", "show matrix stats", "display memory contents", "view matrix status", "show failures", "list warnings", "export matrix data", "backup matrix", "export solutions", or needs to see or export Matrix memory.
This skill should be used when the user asks to "nuke dead code", "find unused exports", "clean up imports", "find orphaned files", "detect circular deps", "cleanup codebase", "find unused packages", "remove dead code", "nuke this", "nuke unused", or needs comprehensive codebase hygiene analysis.
This skill should be used when the user asks to "reindex the codebase", "refresh code index", "update matrix index", "rebuild symbol index", or needs to manually trigger code indexing.
This skill should be used when the user asks to "review this code", "review PR", "code review", "review staged changes", "blast radius analysis", "check impact of changes", or needs comprehensive context-aware code review.
This skill should be used when the user is working with Swift or SwiftUI, asks to "write SwiftUI code", "create iOS app", "fix SwiftUI view", "help with Swift", "SwiftUI architecture", "iOS development", or needs expert-level Swift/SwiftUI guidance with modern patterns and best practices.
This skill should be used when the user asks to "add warning", "remove warning", "list warnings", "check file warning", "check package warning", "manage grudges", or needs to manage file and package warnings in Matrix.
Admin access level
Server config contains admin-level keywords
Executes bash commands
Hook triggers when Bash tool is used
Share bugs, ideas, or general feedback.
Community plugin for Claude Code. Not affiliated with Anthropic.
If Matrix saved you time, be kind and leave a star.
Matrix gives Claude Code persistent memory and deep code intelligence. Solutions survive across sessions. A tree-sitter index makes symbol navigation instant across 15 languages. Hooks catch problems before they happen.
/plugin marketplace add ojowwalker77/Claude-Matrix
/plugin install matrix@ojowwalker77-Claude-Matrix
Requires Bun v1.0+ and Claude Code v2.0+. Verify with /matrix:doctor.
macOS and Linux only. Windows users need WSL.
| Feature | What It Does |
|---|---|
| Memory | Solutions persist. Claude recalls what worked before. |
| Code Index | Find definitions, callers, exports instantly (15 languages) |
| Hooks | Auto-approve reads, catch bad packages, warn on sensitive files |
| Warnings | Track problematic files and packages with CVE checks |
| Code Review | 5-phase analysis with blast radius and impact mapping |
| Deep Research | Multi-source research aggregation |
| Nuke | Dead code, orphaned files, circular dependencies, stale TODOs |
| Context7 | Always-current library documentation |
Solutions persist across sessions with semantic search:
You solve a problem -> Matrix stores it
Similar problem later -> Matrix recalls it
Feedback -> Rankings improve
Tools: matrix_recall matrix_store matrix_reward matrix_failure matrix_status
Fast symbol navigation powered by tree-sitter. Auto-indexed on session start. Respects .gitignore. Uses content hashing for reliable incremental updates.
Languages: TypeScript, JavaScript, Python, Go, Rust, Java, Kotlin, Swift, C#, Ruby, PHP, C, C++, Elixir, Zig
Tools: matrix_find_definition matrix_find_callers matrix_search_symbols matrix_list_exports matrix_get_imports matrix_find_dead_code matrix_find_circular_deps
Matrix runs automatically in the background:
| Trigger | Action |
|---|---|
| Session starts | Index code, initialize memory |
| Tool permission requested | Auto-approve read-only tools |
Before npm install | Check CVEs, deprecation, bundle size |
| Before reading file | Warn if sensitive (.env, keys, secrets) |
| Before editing file | Warn if file has known issues |
| Command | Purpose |
|---|---|
/matrix:review | 5-phase code review with impact analysis |
/matrix:deep-research | Multi-source research aggregation |
/matrix:nuke | Codebase hygiene analysis |
/matrix:doctor | Diagnostics and auto-fix |
/matrix:list | View solutions, stats, warnings |
/matrix:warn | Manage file/package warnings |
/matrix:reindex | Rebuild code index |
Config at ~/.claude/matrix/matrix.config:
{
"hooks": {
"verbosity": "compact",
"permissions": { "autoApproveReadOnly": true }
},
"delegation": { "enabled": true, "model": "haiku" }
}
Verbosity: full (~500 tokens) or compact (~80, recommended)
Model Delegation: Routes simple ops to Haiku for ~40-50% cost savings.
~/.claude/matrix/
├── matrix.db # SQLite database
├── matrix.config # Configuration
├── models/ # Embedding model (~23MB)
└── grammars/ # Tree-sitter parsers
All data local. No external API calls for memory.
Persistent memory system for AI coding sessions — cross-tool memory sharing with 6-dimensional hybrid search
A powerful code indexing tool with multi-platform support
Structural codebase indexing for efficient navigation. Reduces token consumption by 60-80% through targeted line-range reads instead of full file scans.
Self-evolving Claude Code system that learns from corrections, manages context, and improves every session
Optimized file search, semantic indexing, and persistent memory for Claude Code — with optional sync to a self-hosted web dashboard
CodeAlive context engine for semantic code search and AI-powered codebase Q&A. Enables AI coding agents to understand entire codebases beyond just open files — search across all indexed repositories, trace cross-service dependencies, discover usage patterns, and get synthesized answers to architectural questions. Includes a lightweight code exploration subagent, authentication hooks, and multiple search modes (fast lexical, semantic, and deep cross-cutting). Works standalone or alongside the CodeAlive MCP server for direct tool access via the Model Context Protocol.
Modifies files
Hook triggers on file write and edit operations
Modifies files
Hook triggers on file write and edit operations
Share bugs, ideas, or general feedback.