Memstack

MemStack™

[](CHANGELOG.md)

The structured skill framework for Claude Code — 130 professional skills for deployment, security, databases, content, marketing, and more.

Skills activate automatically when you need them. Say "deploy this to Railway" and the right skill loads on demand.

Install

MemStack installs in two parts: the skills (via the Claude Code plugin marketplace) and the engine (the MCP skill loader, via PyPI). You need both — the loader reads skill files from the installed marketplace plugin. Every command below is labeled (in Claude Code) or (in terminal) — running one in the wrong place is the most common setup mistake.

Step 1 — Install the free skills · (in Claude Code):

/plugin marketplace add cwinvestments/memstack
/plugin install memstack@cwinvestments-memstack

Run both commands. This unlocks the 86 free skills right away.

> SSH error? ("Host key verification failed" on a fresh server that's never used GitHub over SSH.) > Default fix · (in terminal) — rewrite GitHub to HTTPS, then retry Step 1: > ``bash > git config --global url."https://github.com/".insteadOf "git@github.com:" > ` > **Backup fix · (in terminal)** — add GitHub's host key, then retry Step 1: > `bash > mkdir -p ~/.ssh && ssh-keyscan github.com >> ~/.ssh/known_hosts > ``

Step 2 — Install the engine · (in terminal):

Windows:

pip install memstack-skill-loader

Linux / Mac — use the explicit interpreter (the same one you register in Step 3):

/usr/bin/python3 -m pip install memstack-skill-loader --break-system-packages

Find yours with which python3 and substitute it if it isn't /usr/bin/python3. The --break-system-packages flag is required on newer externally-managed Python. No pip for that interpreter? Run sudo apt install python3-pip first (Debian/Ubuntu).

Step 3 — Register the MCP server · (in terminal): register against the same interpreter you installed onto in Step 2.

Windows:

claude mcp add --scope user memstack-skills -- python -m memstack_skill_loader

Linux / Mac:

claude mcp add --scope user memstack-skills -- /usr/bin/python3 -m memstack_skill_loader

If this doesn't match Step 2's interpreter, the server won't launch (you'll see a "failed to reconnect" error and activate_license will be missing).

Step 4 — Activate your license · (in Claude Code, after a full restart): fully quit and reopen Claude Code first so it picks up the new MCP server, then run:

activate_license(key="your-key", email="you@example.com")

This unlocks the 44 Pro-exclusive skills (86 free + 44 Pro = 130 total). Free-tier users can skip this step — type list skills to verify the 86 free skills loaded.

See [GETTING-STARTED.md](GETTING-STARTED.md) for detailed setup, and the [Troubleshooting](#troubleshooting) section below.

Troubleshooting

error: externally-managed-environment (during Step 2) — pip refuses to install on a system-managed Python. Add --break-system-packages · (in terminal):

/usr/bin/python3 -m pip install memstack-skill-loader --break-system-packages

Plugin clone fails: "Host key verification failed" (during Step 1) — a fresh server that's never connected to GitHub over SSH. Default fix · (in terminal): git config --global url."https://github.com/".insteadOf "git@github.com:" then retry Step 1. Backup fix · (in terminal): mkdir -p ~/.ssh && ssh-keyscan github.com >> ~/.ssh/known_hosts then retry Step 1.

activate_license not found, or MCP "failed to reconnect (-32000)" — the loader was installed onto a different Python interpreter than the one Claude Code launches (the classic bare-python mismatch on Linux/Mac). Check · (in terminal):

/usr/bin/python3 -c "import memstack_skill_loader; print('ok')"

If it prints ok, make sure Step 3 registered that exact path, then fully restart Claude Code. If it errors with ModuleNotFoundError, reinstall onto the explicit interpreter and re-register against that same path · (in terminal):

/usr/bin/python3 -m pip install memstack-skill-loader --break-system-packages
claude mcp add --scope user memstack-skills -- /usr/bin/python3 -m memstack_skill_loader

Then fully restart Claude Code. Rule of thumb: whatever interpreter you import-check as ok is the path that must appear in your claude mcp add command.

No module named pip (during Step 2) — that interpreter ships without pip (common on minimal Debian/Ubuntu server images). Install it, then retry the Step 2 install · (in terminal):

sudo apt install python3-pip

Tier Structure

| Tier | Skills | Access | |------|--------|--------| | Free | 86 skills | Included with MemStack™ base | | Pro | 130 total (86 free + 44 Pro-exclusive) | Requires Pro Skill Loader (memstack.pro) |

Architecture note: Pro skills are license-gated — when you activate a valid Pro key, the loader downloads them from our server to ~/.memstack/pro-skills. Free users see free skills only; Pro license holders unlock the full catalog. This design keeps a single source codebase with no separate repos or branches for Pro content.

Pro-exclusive skills (44): advanced-security, api-docs, api-load-tester, branching, browser-use, burn, checkpoint, claude-api-helper, codebase-index, config-audit, consolidate, context-db, council, database-architect, database-migration, dependency-auditor, developer-growth-analysis, diagram-generator, doc-index, drift-detection, env-manager-pro, error-handler, frontend-design, git-worktrees, governor-pro, gtm-validator, hooks-integration, ios-app-store, log-analyzer, mcp-builder, meeting-insights-analyzer, model-router, multi-agent, nextjs-conventions, performance-profiler, python-conventions, rag-builder, session-restore, social-media, test-generator, us-privacy-compliance, video-pipeline, video-review, web-scraper — these require an active Pro license.

New skill rule: All newly added skills default to Pro-exclusive. After 90 days, they drop to the free tier unless marked permanent-Pro.

Unlock Pro Skills

Complete the Install steps above first (marketplace + engine). Then, in Claude Code, ask Claude to run:

activate_license(key="your-key", email="you@example.com")

Pro skills download automatically from our server to ~/.memstack/pro-skills — there is no separate marketplace step for Pro. Your key is saved permanently — no environment variables needed.

> Alternative: You can also set MEMSTACK_PRO_LICENSE_KEY as an environment variable instead of using activate_license. Use setx on Windows or add to ~/.bashrc on Mac/Linux, then restart your terminal and Claude Code.

Updating

MemStack updates through three independent channels — refresh whichever you need, then restart Claude Code:

1. Free skills (including newly released ones) — in Claude Code:

/plugin marketplace update cwinvestments-memstack

then /reload-plugins (or restart Claude Code).

Fallback: if a new skill doesn't appear (marketplace version-detection can be unreliable), re-run /plugin install memstack@cwinvestments-memstack. As a last resort, clear the cached plugin and reinstall it.

2. Pro skills — auto-update within 24 hours. To force an immediate re-download, ask Claude to run the refresh_pro_skills tool.

3. Engine (the skill loader itself) — in your terminal:

pip install --upgrade memstack-skill-loader

Free Base (Included)

Everything from MemStack free:

• SQLite-backed persistent memory with semantic vector search
• Deterministic hooks (commit safety, build verification, secret scanning)
• 17 core skills (Echo, Diary, Work, Forge, Scan, Governor, etc.)
• TTS voice notifications (task complete, needs attention, error alerts)
• TokenStack™ context compression
• Always-on rules and slash commands

All Skills (130 total — 86 free + 44 Pro-exclusive)

All 130 skills are fully implemented with complete protocols, context guards, activation messages, and level history. Skills load on-demand via the MCP catalog system — only the skill matching your current task is loaded, preventing context bloat. 44 Pro-exclusive skills require an active license key. Get a key at memstack.pro.

Core (16 — 7 free + 9 Pro)

| Skill | Description | |-------|-------------| | compress | Monitor and manage TokenStack™ context compression for Claude Code sessions, tracking proxy status and token savings. | | diary | Save a structured session diary capturing what was built, decisions made, and next steps at the end of a productive session. | | echo | Recall information from past Claude Code sessions using semantic vector search when referencing previous work. | | goal | Structure a task into TASK, WHY, OUTCOME, CONSTRAINTS, and a required VERIFICATION criterion before execution begins, so 'done' is defined and provable up front. | | grimoire | Manage and update CLAUDE.md files across all projects after significant changes. | | sight | Generate Mermaid diagrams showing project architecture, schema, and data flow for a visual overview of code structure. | | token-optimization | Enable and understand TokenStack™, the built-in compression proxy that shrinks Claude Code tool output before it reaches the Anthropic API, covering how to turn it on, the free and Pro transforms, and how to read token savings on the dashboard. | | burn [PRO] | Track Claude Code token consumption, API spend, and context-window usage with per-session logs, per-project rollups, and monthly budget alerts. | | checkpoint [PRO] | Capture state and context before risky changes so you can roll back if things go wrong, creating structured save points with decision context and file states. | | consolidate [PRO] | Compress a week of diary entries into actionable insights and cross-project pattern summaries. | | context-db [PRO] | Store and query structured project knowledge in a per-project SQLite database, reducing token usage by returning only relevant facts instead of reading full CLAUDE.md. | | council [PRO] | Run a structured multi-perspective debate with four voices analyzing tradeoffs before making significant technical decisions. | | governor-pro [PRO] | Enforce discipline guardrails for agentic coding: think before coding, no sycophancy, no infinite loops, verify before claiming done. | | model-router [PRO] | Route tasks to the right AI model based on complexity, cost, and speed requirements, saving money without sacrificing quality where it matters. | | multi-agent [PRO] | Orchestrate multiple Claude Code instances as a coordinated team where a Manager delegates, a Builder implements, and a Reviewer verifies. | | session-restore [PRO] | Create structured snapshots capturing decisions, context, and next actions for seamless cross-session continuity and handoffs. |

Security (12 — 8 free + 4 Pro)

| Skill | Description | |-------|-------------| | api-audit | Audit Next.js API routes for authentication, authorization, input validation, and common vulnerabilities with a prioritized findings report. | | csp-headers | Audit existing security headers, identify overly permissive directives, and generate a production-ready Content-Security-Policy with companion headers. | | dependency-audit | Scan project dependencies for vulnerabilities, outdated packages, abandoned libraries, and supply chain risks with a prioritized upgrade plan. | | git-guard | Install or verify machine-wide secret-blocking on a repo by confirming the global pre-commit hook, .gitignore coverage, a non-neutered gitleaks config, and a reachable gitleaks binary. | | owasp-top10 | Audit a web application against the OWASP Top 10 (2021) vulnerability categories with actionable findings and remediation steps. | | rls-checker | Audit Supabase Row Level Security policies across all tables in a project to verify table-level access control. | | rls-guardian | Enforce Row Level Security on every Supabase table creation, ensuring no table goes live without proper access policies, with migration templates and pre-commit checks. | | secrets-scanner | Scan a codebase for hardcoded secrets, leaked API keys, and credential exposure across files and git history. | | advanced-security [PRO] | Perform deep security analysis covering authentication, authorization, input validation, API security, dependency CVEs, and infrastructure hardening with proof-of-concept examples and remediation guides. | | config-audit [PRO] | Scan Claude Code project configuration (.claude/, hooks, MCP servers, .gitignore, .env) for secrets, dangerous patterns, and misconfigurations before they cause problems. | | dependency-auditor [PRO] | Perform deep dependency analysis covering CVE vulnerabilities, license compatibility, unused packages, version freshness, and safe update paths. | | env-manager-pro [PRO] | Sync .env files with code, manage multi-environment configs, detect secrets in git, handle rotation workflows, and generate env documentation. |

Deployment (8 — 7 free + 1 Pro)

| Skill | Description | |-------|-------------| | ci-cd-pipeline | Detect the project type and generate a complete CI/CD pipeline with lint, test, build, and deploy stages plus rollback strategy and environment management. | | docker-setup | Analyze a project and generate optimized Dockerfiles, docker-compose configs, and deployment-ready container setups with health checks and volume management. | | domain-ssl | Validate DNS records, SSL certificates, redirects, HSTS, and domain health across all managed properties. | | hetzner-setup | Provision a Hetzner Cloud server with security hardening, reverse proxy, SSL, database setup, monitoring, and automated backups. | | marketplace-submit | Submit skills, plugins, or tools to community marketplaces via pull request with a step-by-step guide. | | netlify-deploy | Validate build config, redirects, environment variables, and deployment readiness for Netlify static/SPA hosting. | | railway-deploy | Validate project configuration, environment variables, and deployment readiness before pushing a Node.js, Python, or Docker application to Railway. | | ios-app-store [PRO] | Follow a step-by-step checklist for submitting apps to the Apple App Store, covering pre-submission, common rejections, Capacitor/React Native specifics, and metadata optimization. |

Development (37 — 16 free + 21 Pro)

| Skill | Description | |-------|-------------| | api-designer | Produce production-ready Next.js App Router API routes with auth guards, Zod validation, typed responses, and consistent error handling. | | changelog-generator | Generate a formatted CHANGELOG.md from git commit history, grouped by type and ready for release. | | code-reviewer | Conduct a systematic code review across security, performance, maintainability, error handling, testing, and accessibility with severity-ranked findings and specific fixes. | | familiar | Break large tasks into coordinated Claude Code session prompts for parallel execution across multiple instances. | | forge | Create new MemStack skills or improve existing ones with proper structure and progressive disclosure. | | mentor | Narrate decisions, tradeoffs, and reasoning in plain language as you build, so the user learns by working alongside you. | | migration-planner | Plan safe database schema migrations with zero-downtime strategies, rollback procedures, data validation checkpoints, and version tracking. | | performance-audit | Identify and prioritize performance bottlenecks across frontend, backend, and network layers with measured impact and fix priority. | | project | Save a

…

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: cwinvestments
• Source: cwinvestments/memstack
• License: MIT
• Homepage: https://memstack.pro

Install and usage instructions live in the source repository linked above.