m2ai-skills-pack

M2AI Skills Pack

A curated Claude Code plugin containing 25 portable skills for strategy work, prompt engineering, model routing, agent auditing, and workflow tooling.

Install

In Claude Code:

/plugin marketplace add m2ai-portfolio/m2ai-skills-pack
/plugin install m2ai-skills-pack@m2ai-skills-pack

Then restart Claude Code. Skills will appear in your skill list and auto-trigger based on their descriptions, or you can invoke them explicitly (e.g. /model-router, /prompt-rewriter).

What's Included

Strategy & Analysis

• executive-briefing — turn a complex event into a structured executive briefing
• counterargument-stress-test — generate and address the N strongest counterarguments to any thesis
• geopolitical-signal-enricher — add geopolitical context to market/tech signals
• bitter-lesson-scorecard — score agent designs against Sutton's Bitter Lesson
• failure-postmortem — structured AI system failure post-mortems
• failure-asymmetry — compare human vs agent invocation behavior of a skill
• aar — after-action review for agent dispatch runs: forensics, failure-mode classification, benchmarks, structured report
• dark-code-audit — scan a repo for modules where human comprehension has gone missing; tiered darkness score with drill-down TODOs

Prompt & Model Engineering

• model-router — classify a task and pick Opus/Sonnet/Haiku with cost delta
• prompt-rewriter — rewrite system prompts stripping compensating complexity
• compensating-complexity-auditor — find scaffolding built around old model limits
• optimize-description — rewrite skill descriptions for discoverability
• spec-gap-detector — stress-test agent specs for ambiguity and missing constraints

Agent & Skill Tooling

• agent-cost-model — model per-task costs, monthly burn, routing savings
• token-burn-auditor — measure session startup overhead and system prompt bloat
• boot-tax-monitor — alert when session startup tokens exceed a threshold
• skill-audit — surface candidates for new skills from conversation patterns
• skill-maintenance — audit skills for quality against Anthropic best practices
• context-hygiene — reference for /by-the-way, /fork, context:fork
• context-fork-guide — add context:fork to heavy-research skills

Workflow

• gh-review — review GitHub repos against current project, generate HTML report
• get-api-docs — fetch API docs via Context Hub (/chub) to verify model names
• file-intel — Gemini-powered extraction/summary for PDF/PPTX/XLSX/DOCX/CSV folders
• banana-maker — Gemini image generation via Nano Banana Pro prompting
• what-am-i-forgetting — consolidated agenda recall across memory index, roadmaps, open-item queues, daily notes, crons, and project folders

Configuration

Some skills expect environment variables for user-specific paths. If unset, each skill falls back gracefully and flags the missing source in its output.

Variable	Used by	Purpose
$PROJECTS_DIR	what-am-i-forgetting	Root of your code projects (e.g. ~/projects)
$VAULT_DIR	what-am-i-forgetting, aar	Notes vault for daily notes and saved reports
$MEMORY_INDEX	what-am-i-forgetting	Path to your memory-index file
$MEMORY_DIR	what-am-i-forgetting	Directory containing memory files
$ORCHESTRATOR_DB	aar	SQLite DB holding agent mission/task state
$OPEN_ITEMS_JSON	aar	Queue file for action-item append
GEMINI_API_KEY	file-intel, banana-maker	Gemini API access

Known Rough Edges

A few skills contain environment-specific references. They still work, but you may want to adjust:

• gh-review — hardcoded LAN IP 10.0.0.46 for the report-server URL. Swap for your own file server, or ignore (the HTML file is still generated locally).
• banana-maker — one mention of a "Surface tablet." Harmless.
• context-fork-guide — example command points at /home/apexaipc/.claude/skills/. Replace with your skills path if you run the example verbatim.
• get-api-docs — depends on the /chub (Context Hub) skill being installed separately. Without it, this skill is a no-op.
• aar — assumes an orchestrated agent system with a SQLite task DB and A2A endpoints. Without those, Phase 1 forensics gets thin.
• what-am-i-forgetting — most useful if you maintain a memory-index file, per-project TODO/NEXT files, and daily notes. Without them, it falls back to folder inventory + cron listings.

Version

0.2.0 — adds aar, dark-code-audit, what-am-i-forgetting. Feedback welcome.