6 free skills for Claude Code. Safety guardrails, structured debugging, product stress-testing, skill optimization, code principles, and AI co-creation philosophy.
Built by Menteorama Studio — an AI integration studio that ships with Claude Code every day.
claude plugin marketplace add menteorama/essentials --sparse .claude-plugin
claude plugin install menteorama-essentials@menteorama-essentialsTwo commands. Six skills, ready to use.
Two hooks that protect you from three categories of disaster:
Command protection — Catches rm -rf, DROP TABLE, git push --force, git reset --hard, git clean -f, curl | bash, chmod 777, accidental npm publish, and more. Allows safe deletions (node_modules, .next, dist) automatically.
File protection — Warns before editing .env files, SSH keys, certificates, credentials, .git/ internals, and anything in secrets/ directories.
Secrets detection — Scans both commands and file content for AWS keys, GitHub tokens, API keys (sk-*), Slack tokens, private key blocks, and database connection strings with embedded passwords.
How it works: Two PreToolUse hooks — one for Bash commands, one for Write/Edit operations. No config needed — they just work.
A Staff Debugger protocol. Four phases: Investigate, Analyze, Hypothesize, Implement. The iron law: no fixes without root cause. Stops Claude from applying band-aids and forces it to actually understand what's broken before touching code.
Trigger it: "debug this", "why is this broken", "fix this bug"
Six forcing questions that kill bad ideas before you waste months building them. Demand reality, status quo analysis, desperate specificity, narrowest wedge, unique observation, AI future-fit. Produces a BUILD / ITERATE / KILL verdict.
Trigger it: "I have an idea", "is this worth building", "office hours"
Architecture first, code second. Opinionated defaults. Error paths as first-class citizens. Premise challenges before implementation. Loads automatically when you ask Claude to build something.
Trigger it: "build this", "architect this", "how should I build"
Turns Claude from an assistant into a peer. Challenge premises, strategic depth first, ship with momentum, ground decisions in meaning. The operating system for thinking with AI, not just using it.
Trigger it: "think with me", "co-create", "second brain mode"
Analyzes your installed skills, maps them to 7 task types (build, debug, deploy, audit, design, strategy, memory), and generates a CLAUDE.md routing rule. Based on Meta-Harness research: targeted skill loading beats trigger-based activation by +51 points in benchmarks.
Loading 3-4 targeted skills per task type beats loading 16+ via triggers. Less context = better outcomes.
Trigger it: "optimize my setup", "skill audit", "too many skills"
We use 62 skills daily at Menteorama Studio. These are the six we couldn't work without — the ones that changed how we interact with Claude Code:
- careful saved us from a
DROP TABLEin production - investigate cut our debugging time in half
- office-hours killed 3 ideas that would have wasted months
- code eliminated "it depends" from Claude's vocabulary
- core turned Claude into a co-founder, not a code monkey
- optimize proved we were loading too many skills — and fixed it
These 6 skills are the free layer of the Menteorama Brain Kit — a complete Claude Code second brain with 62 skills across 9 categories:
| Category | Skills | Examples |
|---|---|---|
| Marketing | 12 | SEO, content strategy, cold email, ad creative |
| Design | 4 | Brand systems, UI/UX, logo generation, slides |
| Strategy | 6 | Narrator (brand storytelling), pricing, launch |
| Sales | 5 | Funnels, proposals, enablement, RevOps |
| CRO | 7 | Signup flows, paywalls, popups, onboarding |
| Research | 3 | Customer research, competitor analysis |
| AI | 2 | AI-native SEO, token optimization |
| Development | 7 | Frontend design, Figma bridge, web audits |
| Methodology | 10 | Superpowers (agentic dev), Meta-Harness, GSD, retros |
Install the full Brain Kit:
/bin/bash -c "$(curl -fsSL https://lab.menteorama.co/install.sh)"One command. No prerequisites. Works on Mac and Linux. Bilingual (Spanish/English).
Menteorama Studio builds AI-powered tools for entrepreneurs, creators, and operators who think with AI, not just use it. We're based in Mexico and serve the Spanish-speaking market — 650M people with almost zero Claude Code adoption. Yet.
MIT