Curated directory

AI Resources

The signal, without the noise. A short, curated set of the best places to stay informed, get credentialed and start building, chosen for executives who are early in their AI journey. Quality over quantity, always.

Stay informed

The mistake is subscribing to everything. The daily newsletters cover the same launches, so more of them is not more signal. Pick one daily briefing to know what happened, and one weekly source to understand what it means. That is the whole system.

Daily briefings — pick one

  • The Rundown AI

    Rowan Cheung

    Daily AI news, tools and practical insight in plain language. The strongest default daily for a non-technical leader.

  • The Neuron

    Pete Huang and Noah Edelman

    Daily AI news plus actionable guidance, written in a friendly, easy-to-read tone. A lighter alternative to The Rundown.

  • Superhuman AI

    Zain Kahn

    Get smarter about AI in three minutes a day: tools, tutorials and the business angle. The shortest serious daily brief.

Weekly depth — pick one or two

  • One Useful Thing

    Prof. Ethan Mollick, Wharton

    Research-based essays on what AI means for work, management and decision-making. The single best fit for this audience.

  • The Batch

    Andrew Ng, DeepLearning.AI

    A weekly, hype-free roundup of AI news and research with commentary from one of the most credible voices in the field.

  • TLDR AI

    The TLDR newsletter family

    A technical daily covering research, engineering and big-tech AI news. Substance over hype, for the more technical leader.

  • Ben's Bites

    Ben Tossell

    AI news through a founder, builder and early-investor lens. Best for leaders with a startup or product orientation.

Creators and specialists — follow as needed

  • Nate Herk

    YouTube and the AI Automation Society

    No-code AI automation and agents, built on n8n. The most builder-oriented resource here: learn to actually build working automations.

  • Eric Vyacheslav

    LinkedIn

    A leading AI voice on LinkedIn: a steady, digestible feed of model releases and breakthroughs inside a platform you already use.

  • Guillermo Flor

    The AI Opportunity

    AI as a business and investment opportunity: where the money is moving, for leaders who think in markets, not tutorials.

  • John Peslar

    LinkedIn and newsletter

    Hands-on AI automation tactics for sales, lead generation and personal-brand growth on LinkedIn.

  • AlphaSignal

    Technical AI newsletter

    Research papers, model releases and trending open-source tools, built to read in five minutes. For the technically curious leader.

  • There's An AI For That (TAAFT)

    AI tools directory

    A curated, searchable catalog of thousands of AI tools by use case. The place to go in "what tool do I use for X" mode.

Get certified

Courses and certifications worth your time, grouped by function. Most of the best ones are free; the few paid credentials are marked. Course details and prices change, so confirm on the provider page before you enroll.

AI literacy for leaders

Start here. Broad, non-technical grounding in what AI is and how to use it at work.

  • Generative AI for EveryoneFree to audit · ~$49 for the certificate

    DeepLearning.AI (Andrew Ng)

    The flagship non-technical primer. Best first course for any executive.

  • AI Fluency: Framework and FoundationsFree · includes a certificate

    Anthropic Academy

    A vendor-neutral, ethics-aware framework for working with any AI tool.

  • OpenAI Academy

    A self-paced primer you work through inside ChatGPT itself.

  • Google Cloud Generative AI LeaderPaid · ~$99 proctored exam (prep free)

    Google Cloud

    A real proctored certification designed for non-technical leaders.

Prompt engineering

The core practical skill: getting consistently strong output from any AI model.

AI for product and business

Planning AI initiatives, organizational readiness and AI strategy.

Technical and machine learning

For leaders who want genuine fundamentals, or a credential their teams can pursue.

  • Machine Learning SpecializationFree to audit · ~$49/month for the certificate

    Stanford and DeepLearning.AI

    The canonical on-ramp to machine learning fundamentals.

  • AWS Certified AI PractitionerPaid · ~$100 proctored exam

    Amazon Web Services

    A foundational, application-focused credential accessible to non-engineers.

  • Microsoft

    Entry-level technical credential. Note: the older AI-900 retires mid-2026; pursue AI-901.

AI governance, risk and ethics

For leaders in legal, compliance and risk, or anyone building an AI governance program.

The toolkit

Open-source repositories worth knowing as you move from learning to building: the actual toolkit behind a power-user Claude Code setup, grouped by what each one does for you. Every entry below carries four lenses so you can decide what is worth installing.

  • Quality: what it does for the build itself.
  • Tokens: the cost lens. Higher, lower or neutral. Percentages only when a repo states them.
  • Commercial: where the money is if you operate it for clients.
  • How to use: the first command, the first decision, the first link.

Claude Code, supercharged

Power-user setups that make Claude Code itself dramatically more capable. Install these first.

  • superpowers/obra·Auto-sources and activates skills
    Quality
    High. Skills load on demand, so Claude reaches for the right capability without you naming it.
    Tokens
    Lower per-task: only the active skill loads, instead of a fat system prompt.
    Commercial
    Resell your own skill packs to clients or teams.
    How to use
    `git clone` and follow the README to point Claude at the skills directory.
  • everything-claude-code/affaan-m·The complete Claude Code playbook
    Quality
    High. A curated reference set of hooks, plans and patterns.
    Tokens
    Neutral. Reference material you read, not load at runtime.
    Commercial
    A reading list, not a product.
    How to use
    Browse the GitHub README. Cherry-pick the techniques that apply to your stack.
  • gstack/garrytan·A full deploy pipeline inside Claude
    Quality
    High. Turns Claude into a build/test/deploy operator with real guardrails.
    Tokens
    Neutral. Token cost trades against the human-hours it removes.
    Commercial
    Strong: ship client work end-to-end from a single chat.
    How to use
    Install per the README. Point it at a git repo with deploy targets configured.
  • gbrain/garrytan·A persistent second brain for Claude
    Quality
    High. Long-running context survives sessions and projects.
    Tokens
    Lower over time: retrieves only relevant memory instead of replaying full history.
    Commercial
    Differentiator for client work: continuity across engagements.
    How to use
    Install, point it at the projects you want a brain for, let it index.
  • Personal_AI_Infrastructure/danielmiessler·A full personal AI operating system
    Quality
    High. The opinionated reference architecture for running AI on your own stack.
    Tokens
    Variable: depends on which modules you enable.
    Commercial
    A great starting point for productizing an internal AI ops practice.
    How to use
    Read the README first; this is a system, not a single tool. Adopt module by module.
  • claude-doctor/millionco·Diagnoses and tunes your setup
    Quality
    High for reliability. Catches silent misconfigurations.
    Tokens
    Lower indirectly: fewer wasted retries from broken setups.
    Commercial
    Run it for clients as a paid setup audit.
    How to use
    Run the doctor command, follow its fixes.
  • archon/coleam00·Repeatable, standardized coding workflows
    Quality
    High. Codifies "the right way" to do common engineering tasks.
    Tokens
    Lower per task: shorter prompts, standardized scaffolds.
    Commercial
    Productize your engineering playbook for client teams.
    How to use
    Clone, define your workflows in the prescribed format, invoke from Claude.

Memory and token efficiency

The single highest-leverage category for cost. Memory and compaction repos cut what you pay for repeated context.

  • claude-mem/thedotmack·Free long-term memory, fewer tokens
    Quality
    High. Claude stops forgetting between sessions on the same project.
    Tokens
    Materially lower: retrieves the relevant slice instead of replaying full history. Cost scales with what you remember, not how often you ask.
    Commercial
    Premium feature for client engagements that span weeks.
    How to use
    Install per README, give it a project path, let it manage the memory store.
  • mempalace/MemPalace·A structured memory palace
    Quality
    High when your work is reference-heavy.
    Tokens
    Lower: structured retrieval beats raw context dumps.
    Commercial
    Sell as a knowledge layer for client teams.
    How to use
    Install per README. Ingest your domain documents, then query.
  • rtk/rtk-ai·Aggressive prompt compaction
    Quality
    Neutral to slightly improved: tighter prompts, less drift.
    Tokens
    The project advertises up to 90% reduction on suitable workloads. Measure on your own corpus before trusting any figure on yours.
    Commercial
    Real margin on high-volume agent or pipeline work.
    How to use
    Install per README, wrap your prompt pipeline with the rtk client.
  • lat.md/1st1·Fixes markdown file architecture
    Quality
    Higher: cleaner files, better parseability for AI agents.
    Tokens
    Lower indirectly: tidier files mean smaller, focused reads.
    Commercial
    Niche, but real if you sell content systems.
    How to use
    Install and run against your markdown tree.

Agents and skills

Frameworks and patterns for turning Claude into a fleet of specialists, not a single chat.

  • agent-skills/addyosmani·A curated agent-skills collection
    Quality
    High. Vetted skills you do not have to invent.
    Tokens
    Lower per task: load only the skill the moment needs.
    Commercial
    Use these as starter blocks for client agents.
    How to use
    Browse the catalog, copy the skill folder, point Claude at it.
  • book-to-skill/virgiliojr94·Turns any book into a Claude skill
    Quality
    High for domain expertise. Embeds the book's framework into Claude.
    Tokens
    Lower than copy-pasting the book; higher than no context. Net positive.
    Commercial
    Productize your own book as a paid skill. Direct monetization.
    How to use
    Feed it a PDF or text. Out comes a skill folder you ship.
  • create-agent-tui/OpenRouterTeam·A full agent in one command
    Quality
    High starter quality. Right defaults, fewer footguns.
    Tokens
    Neutral. Scaffolds the agent, you choose the model.
    Commercial
    Fastest path from idea to demoable client agent.
    How to use
    Run the command, answer the TUI prompts, get a working agent.
  • maestro/its-maestro-baby·A Bloomberg Terminal for Claude Code
    Quality
    High visibility into what the agent is doing.
    Tokens
    Neutral to slightly lower: easier to spot wasteful patterns.
    Commercial
    A dashboard you can demo to enterprise buyers.
    How to use
    Install, point at your Claude sessions, watch the panes.
  • helmor/dohooo·Multi-agent development fixer
    Quality
    High on complex bug hunts and refactors.
    Tokens
    Higher per task than a single agent, but lower than failing repeatedly.
    Commercial
    Premium "stuck-codebase rescue" engagements.
    How to use
    Install, point at the broken repo, supervise.
  • paperclip/paperclipai·An AI agent that manages agents
    Quality
    High for orchestration at scale.
    Tokens
    Variable: a meta-agent uses tokens, but it offsets coordination overhead.
    Commercial
    Strong: sell agent-fleet operations to enterprise clients.
    How to use
    Install per README, give it the agents to manage.
  • OpenMythos/kyegomez·An open-source agent toolkit
    Quality
    High when you need a vendor-neutral base.
    Tokens
    Variable by configuration.
    Commercial
    Build proprietary agents without lock-in.
    How to use
    Clone, install dependencies, pick the agent type to extend.
  • token-dashboard/nateherkai·Live Claude token tracking
    Quality
    Neutral. Diagnostic.
    Tokens
    Lower indirectly: you cannot cut what you do not measure.
    Commercial
    Run it for clients to justify AI spend with real numbers.
    How to use
    Install, point at your usage source, watch the live dashboard.

Content, media and design

Ship publishable output (decks, audio, print, design) without paying SaaS rent.

  • PptxGenJS/gitbrent·Generate slide decks programmatically
    Quality
    High when content is data-driven. Pixel-perfect on layouts you control.
    Tokens
    Neutral. The model writes content; this writes the pptx.
    Commercial
    Strong: an alternative to monthly Gamma fees.
    How to use
    `npm install pptxgenjs`. Build a script that takes structured content and emits a deck.
  • supertonic/supertone-inc·Fast TTS, an ElevenLabs alternative
    Quality
    High for usable narration without a subscription.
    Tokens
    N/A. Saves SaaS dollars, not tokens.
    Commercial
    Strong: podcast and video production at lower cost.
    How to use
    Install per README. Feed it text, get audio.
  • LuxTTS/ysharma3501·Free text-to-speech for Claude
    Quality
    Good for quick voiceovers; not a replacement for studio TTS.
    Tokens
    N/A.
    Commercial
    Use as a free fallback when SaaS quotas run out.
    How to use
    Install per README. Plug into your Claude pipeline.
  • cli-printing-press/mvanhorn·Publication-quality output from the CLI
    Quality
    High typographic quality (LaTeX-class).
    Tokens
    Neutral. The press handles formatting; the model writes.
    Commercial
    Sell self-published books and reports without a designer.
    How to use
    Install per README, hand it markdown, get a print-ready PDF.
  • open-design/nexu-io·An open design system for Claude builds
    Quality
    High visual consistency across AI-generated UIs.
    Tokens
    Neutral.
    Commercial
    Use as the house style for productized client builds.
    How to use
    Install per README, reference the design tokens in your Claude prompts and components.

Web, data and research

Give Claude eyes and hands on the open web and your own data.

  • browser-harness/browser-use·Flexible AI browser control
    Quality
    High for any task that needs a real browser.
    Tokens
    Neutral to higher: the browser context is not free; cap rounds.
    Commercial
    Strong: scraping, QA, research, lead gen.
    How to use
    Install per README, give Claude a goal and a URL, watch it work.
  • Scrapling/D4Vinci·A robust, free web scraper
    Quality
    High on hostile sites where naive scrapers fail.
    Tokens
    N/A. Saves time and SaaS costs.
    Commercial
    Strong: an alternative to paid scraping APIs for research and lead gen.
    How to use
    `pip install scrapling`. Build the scraper into your research pipeline.
  • wterm/vercel-labs·A web terminal, by Vercel
    Quality
    High for client-facing terminal experiences.
    Tokens
    Neutral.
    Commercial
    Embed in deliverables that include a live shell.
    How to use
    Install per README, expose the terminal in your app.
  • data-formulator/microsoft·Natural-language BI and charts
    Quality
    High for fast exploratory analysis.
    Tokens
    Variable, depends on dataset size. Cap with sampling.
    Commercial
    Replace a chunk of Tableau or PowerBI usage for clients.
    How to use
    Install per README. Load a dataset, ask in natural language.

Privacy, learning and career

Govern what leaves your environment, accelerate your own skills, and ship things that earn.

  • LLM-anonymization/zeroc00I·Strips sensitive data from prompts
    Quality
    High. Real governance, not vibes.
    Tokens
    Neutral.
    Commercial
    Strong for any enterprise sale: handle PII safely.
    How to use
    Install per README. Wrap your prompt pipeline with the anonymizer.
  • ai-engineering-from-scratch/rohitg00·A 428-lesson AI engineering course
    Quality
    High for technical depth, well beyond the academy.
    Tokens
    N/A.
    Commercial
    Builds the skills behind paid AI engineering work.
    How to use
    Start at the README index. Work through the lessons in order.
  • omniget/tonhowtf·Free courses pulled from everywhere
    Quality
    High coverage, variable depth.
    Tokens
    N/A.
    Commercial
    Cheap upskilling for you or your team.
    How to use
    Install per README, search by topic.
  • career-ops/santifer·An AI-powered job-search system
    Quality
    High for structured search at scale.
    Tokens
    Neutral.
    Commercial
    Operate as a paid service for executive job-seekers.
    How to use
    Install per README, configure your target roles, let it run.