By EugOT
Native Zig local memory for Claude Code with MCP tools, Markdown wiki storage, and session checkpoints.
This plugin requires configuration values that are prompted when the plugin is enabled. Sensitive values are stored in your system keychain.
cortex_homeOptional local data directory. Leave empty to use the Claude plugin data directory or the default home directory.
Matches all tools
Hooks run on every tool call, not just specific ones
Admin access level
Server config contains admin-level keywords
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
${user_config.cortex_home}database_urlDeprecated compatibility setting from the previous PostgreSQL runtime. Native Cortex ignores this value; use Cortex home for local file-backed storage.
${user_config.database_url}Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
Crtx is a local memory server for Claude Code. This branch replaces the former dynamic-runtime implementation with a native Zig command line and MCP server.
The retained native surface is deliberately narrow and evidence-based:
Unsupported historical surfaces remain visible only where client compatibility requires a tool name. They return explicit native unsupported status.
Install Zig 0.16 or newer first; the Claude plugin postInstall step expects
zig on PATH.
zig build
zig build test
zig build run -- doctor
zig build run -- remember --content "Decision: keep Crtx native." --tag decision --force
zig build run -- recall --query "Crtx native" --include-related
zig build run -- stats
zig build run -- wiki write notes/native.md "# Native Crtx"
zig build run -- wiki read notes/native.md
Set CORTEX_HOME to choose the store directory. Without it, Crtx uses
CLAUDE_PLUGIN_DATA when launched as a Claude plugin, otherwise
$HOME/.claude/methodology/native.
The previous database_url plugin setting is deprecated and ignored by the
native runtime. Existing plugin configs can leave it in place during migration,
but new deployments should use cortex_home or CORTEX_HOME.
zig build run -- mcp
The server supports initialize, tools/list, and tools/call.
Implemented core tools include remember, recall, unified_search,
memory_stats, get_telemetry, checkpoint, detect_domain, list_domains,
get_methodology_graph, query_workflow_graph, navigate_memory, and the
Markdown wiki tools.
zig build fmt
zig build check
zig build test
zig build docs
gitleaks detect --source . --config .gitleaks.toml --redact
git ls-files | rg '(\.(py|pyi|ipynb|js|mjs|cjs|jsx|ts|tsx)$|(^|/)(package\.json|pyproject\.toml|setup\.py|requirements\.txt|tox\.ini|pytest\.ini|tsconfig\.json|bun\.lock|package-lock\.json|pnpm-lock\.yaml|yarn\.lock|pixi\.lock|uv\.lock)$)'
The final command must print nothing.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Cavekit 4 — compressed spec-driven dev. Three commands (/ck:spec, /ck:build, /ck:check) over one SPEC.md file. Caveman encoding. Bug-to-spec backprop.
npx claudepluginhub eugot/crtxv9.52.0 - Reliability wave: tangle contextual review correction loop with hard round ceiling, progress-supervised review rounds (per-agent stall watch, descendant-tree kills), council diversity and agy pin fixes, marketplace generator source-of-truth fix, provider troubleshooting runbook and cost-expectations docs. Run /octo:setup.
Comprehensive C4 architecture documentation workflow with bottom-up code analysis, component synthesis, container mapping, and context diagram generation
Upstash Context7 MCP server for up-to-date documentation lookup. Pull version-specific documentation and code examples directly from source repositories into your LLM context.
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
The Claude Code knowledge system — 380+ skills, 182+ agents, 100+ commands, 40 hooks, 32 rules, and workflows.
Comprehensive feature development workflow with specialized agents for codebase exploration, architecture design, and quality review