Best AI Skills for Productivity & Organization in 2026: 15 Compared
Fifteen productivity skills ranked from 454 candidates. The category splits cleanly: meta-cognition (plan/execute) versus office-app integrations (Notion, PowerPoint, Lark). Both win.
Productivity & Organization is the second-biggest category in the catalog (454 skills) and the most internally diverse. The top 15 splits into three almost-equal clusters: meta-cognition (plan-then-execute discipline from obra/superpowers), office-app integrations (Notion, PowerPoint, Lark, Obsidian, Apple Notes — agents reaching INTO the tools people already use), and response compression (skills that make Claude itself shorter and faster). The third group is the one I expect newer users to skip and then regret skipping. We ranked all 454 and picked the top 15.
Quick Pick
executing-plans — paired with #2 (writing-plans), this is the meta-skill that turns Claude Code from a stream-of-consciousness assistant into something resembling actual engineering discipline. Write the plan in one session, execute it with review checkpoints in another. The #1 pick because it changes the shape of every other workflow.
What These Skills Actually Do
Productivity skills sit at the workflow layer — the place where individual tasks meet the broader system of how you actually get work done. The category divides cleanly:
- Planning + execution discipline — writing-plans and executing-plans force a separation between thinking and doing that solo founders especially benefit from
- Office-app integrations — Notion, PowerPoint, Lark Docs, Obsidian, Apple Notes (agents reaching INTO the tools you already use, rather than asking you to switch)
- Token compression — Simple Response Mode, Caveman, Caveman Bundle: skills that change how Claude responds, not what it does
- Automation primitives — Automation Workflows, ClawdHub for skill management
A great productivity skill changes a rhythm, not a task. The mediocre ones automate one specific thing and produce a temporary win. The ones that rank high here all show up dozens of times per week.
How We Ranked
We sorted 15 candidate skills by a composite score:
- Popularity signal — the highest of GitHub stars, install count, or ClawHub download count. Log-scaled so a 100-star skill doesn't get buried under a 100,000-star one if the smaller one is meaningfully better.
- Quality score — when set, a 0–5 rubric that breaks ties within popularity tiers. Roughly 15% of catalog skills carry a quality score today; we surface it in the comparison table when available.
The formula is identical across the entire Best-Of 2026 series, so you can compare apples to apples between categories.
The Top 15
1. executing-plans
Skill · obra/superpowers · 167.5K signal · quality 4/5 Use when you have a written implementation plan to execute in a separate session with review checkpoints.
The take: The mechanic that makes writing-plans useful. Without an execution discipline, written plans become aspirational documents that drift; this skill enforces the review checkpoints that keep execution honest. Pair them; install separately if you only want one half.
2. writing-plans
Skill · obra/superpowers · 167.5K signal · quality 4/5 Use when you have a spec or requirements for a multi-step task, before touching code.
The take: The other half of the plan-then-execute pair. Forces structured thinking before implementation — exactly the discipline missing from "I have an idea, let me build it" workflows that produce the 42% failure mode the Founder's Playbook flagged. Install both #1 and #2 together; they're co-designed.
3. notion
Skill · NousResearch/hermes-agent · 124.8K signal · quality 4/5 Notion API via curl: pages, databases, blocks, search.
The take: Direct API access via curl rather than a heavier SDK wrapper — means it just works in any environment without dependency management. The 4/5 quality score is the highest in the Hermes office-tools cluster. If you live in Notion, install first.
4. powerpoint
Skill · NousResearch/hermes-agent · 124.8K signal · quality 4/5 Create, read, edit .pptx decks, slides, notes, templates.
The take: A different theory of PPTX from the Documentation & Writing #8 pick — this one is part of the Hermes agent runtime, so it composes with the rest of the Hermes skill set. Install this if you're in Hermes; the Documentation pick (Powerpoint / PPTX) if you're standalone.
5. maps
Skill · NousResearch/hermes-agent · 124.8K signal · quality 4/5 Geocode, POIs, routes, timezones via OpenStreetMap/OSRM.
The take: OpenStreetMap/OSRM means no API key, no quota, no Google Cloud bill. Perfect for the niche of "I need geo data in my agent workflow but don't want to provision a Maps API account." More limited than Google Maps for some POI lookups; sufficient for most.
6. Simple Response Mode
Command · roin-orca/skills · 98.7K signal · quality 4/5 Simplifies AI responses to be shorter, clearer, and more direct. Reduces verbosity, cuts unnecessary explanations.
The take: The "stop padding your responses" command. Cuts the meta-commentary ("Great question! Let me think about this...") that bloats every Claude response by 10-30%. Install once, never look back.
7. Caveman
Command · juliusbrussee/caveman · 92.3K signal · quality 4/5 Compress AI responses into telegraphic "caveman speak" — minimal words, maximum signal. Reduces token usage by 30-50% in long sessions.
The take: More aggressive than #6. The output reads weird at first ("Plan good. Execute soon. Tests pass."), but the 30-50% token savings on long sessions is real, measurable money. Install when you're hitting context limits or paying for tokens directly. Skip if you're sharing chat logs with humans.
8. Lark Documents
Plugin · larksuite/cli · 83.3K signal · quality 4/5 Create, edit, and manage Lark/Feishu documents using the official Lark Open API. Official Lark skill.
The take: The "official" matters — built by Lark themselves rather than reverse-engineered. Reliability is meaningfully higher than community-built integrations for proprietary platforms. Required install for anyone working at a Lark-using org (most large Chinese tech, plus a growing US/EU footprint).
9. Lark Approval Workflows
Plugin · larksuite/cli · 56.1K signal · quality 4/5 Automate Lark/Feishu approval workflows: create, submit, query, and manage approval forms programmatically.
The take: The unsexy enterprise primitive. Most internal processes at Lark-using orgs route through approval forms; this skill makes them programmable. Niche outside of Lark; load-bearing inside.
10. Caveman Token Compression Bundle
Command · juliusbrussee/caveman · 50.3K signal · quality 4/5 Caveman bundle: compress AI responses to use fewer tokens with telegraphic "caveman speak". Includes commit, review, compress, and help.
The take: The packaged version of #7 plus three additional Caveman commands. Install this instead of just #7 if you want the full toolkit; install just #7 if you only want response compression in your main flow.
11. Obsidian
Skill · clawhub/obsidian · 72.9K signal · quality unrated Work with Obsidian vaults (plain Markdown notes) and automate via obsidian-cli.
The take: Obsidian users skew strongly toward developers and writers — Claude reaching INTO an Obsidian vault closes a workflow loop. Plain-markdown architecture means Claude doesn't need a heavy plugin layer; it just reads and writes files. The right install if you write in Obsidian and want Claude to surface, summarize, or restructure your notes.
12. Value-First Development
Command · hugmouse/skills · 35.4K signal · quality 3/5 Promotes value-first thinking in development: prioritize user value, defer technical complexity, and ship the simplest solution that works.
The take: A philosophical guardrail more than a tactical tool. Reframes Claude's default tendency toward thorough/complex solutions into "what's the simplest version that actually delivers value?" Pairs well with #1 and #2 — the planning skills constrain what gets built; this one constrains how.
13. Automation Workflows
Skill · clawhub/automation-workflows · 58.5K signal · quality unrated Design and implement automation workflows to save time and scale operations as a solopreneur. Identify repetitive tasks, build workflows across tools.
The take: The "solo founder ops" skill. Less about a specific automation and more about the meta-skill of recognizing which manual tasks are worth automating. Strong for the one-person-company audience.
14. Clawdhub
Skill · steipete/clawdhub · 28.2K signal · quality unrated ClawdHub CLI to search, install, update, and publish agent skills from clawdhub.com.
The take: Meta-skill for managing your skill collection. Useful for "what skills are out of date" and "find me a skill for X" without leaving the terminal. Becomes more valuable the more skills you have installed.
15. Apple Notes
Skill · clawhub/apple-notes · 26.8K signal · quality unrated
Manage Apple Notes via the memo CLI on macOS. Create, view, edit, delete, search, move, and export notes.
The take: macOS-only by definition. The right install if Apple Notes is where you actually capture things during the day — bridges the gap between "I just thought of something" and "Claude knows about it." Most cross-platform users will skip this for Obsidian (#11) or Notion (#3).
Comparison Table
| # | Skill | Type | Stars / Installs | Quality | License |
|---|---|---|---|---|---|
| 1 | executing-plans | Skill | 167.5K | 4/5 | — |
| 2 | writing-plans | Skill | 167.5K | 4/5 | — |
| 3 | notion | Skill | 124.8K | 4/5 | MIT |
| 4 | powerpoint | Skill | 124.8K | 4/5 | Proprietary. LICENSE.txt has complete terms |
| 5 | maps | Skill | 124.8K | 4/5 | MIT |
| 6 | Simple Response Mode | Command | 98.7K | 4/5 | MIT |
| 7 | Caveman | Command | 92.3K | 4/5 | MIT |
| 8 | Lark Documents | Plugin | 83.3K | 4/5 | MIT |
| 9 | Lark Approval Workflows | Plugin | 56.1K | 4/5 | MIT |
| 10 | Caveman Token Compression Bundle | Command | 50.3K | 4/5 | MIT |
| 11 | Obsidian | Skill | 72.9K | — | MIT-0 (Free to use, modify, and redistribute. No a |
| 12 | Value-First Development | Command | 35.4K | 3/5 | MIT |
| 13 | Automation Workflows | Skill | 58.5K | — | MIT-0 (Free to use, modify, and redistribute. No a |
| 14 | Clawdhub | Skill | 28.2K | — | — |
| 15 | Apple Notes | Skill | 26.8K | — | MIT-0 (Free to use, modify, and redistribute. No a |
FAQ
How is this list different from the category page on aiskill.market?
The category page is a directory: every skill in the category, sortable and filterable. This list is editorial — opinionated, time-stamped (2026-05-17), and ranked. Use the directory when you know what you want; use this when you don't.
Why does the #1 pick have fewer stars than #5?
Stars are one signal among several. The composite score above also includes install counts (which reflect actual usage on aiskill.market) and the optional quality score. A skill with a smaller star count can rank higher if its installs or quality score are strong enough to offset.
Are these all free?
Most are. Notion, Maps, Simple Response Mode, Caveman (both #7 and #10), Lark Documents, Lark Approval Workflows, and Value-First Development are explicitly MIT. The PowerPoint skill at #4 has a proprietary license (free to use, but not freely redistributable — see the upstream LICENSE.txt). The Apple Notes skill requires macOS but the skill itself is free.
How do I install one?
Each linked skill page has install instructions. The fastest path is the one-line install via the aiskill.market CLI or by adding the source repo as a Claude Code plugin marketplace.
How often does this list update?
Quarterly. We re-run the generator script against the catalog every three months. Productivity rankings change more often than security or DevOps because new note-taking and workflow tools enter the catalog regularly.
What's the starter pack for someone new to the category?
Five skills: writing-plans (#2), executing-plans (#1), Simple Response Mode (#6), one note system (Notion #3 or Obsidian #11 depending on your tool), and one office integration if needed (PowerPoint, Lark, etc.). That covers the meta-cognition layer + your tool surface + Claude's verbosity. Add more as you encounter the specific gap each fills.
Does Caveman work for code generation too, or just prose?
It's framed as response compression generally but works best on explanatory text. Code generation passes through largely unmodified — which is good, because you don't want compressed code. The savings come from comments, explanations, and the meta-commentary around the code.
Related Categories
- Best AI Skills for Development & Code Tools in 2026
- Best AI Skills for AI & Machine Learning in 2026
- Caveman Token Compression — a longer essay on why response compression is a load-bearing productivity move
Browse The Full Catalog
Find every skill in this category — including the ones that didn't make the top 15 — at the Productivity & Organization page.
Part of the Best-Of 2026 series. Updated 2026-05-17. Skills sampled from a catalog of ~262 active entries with a combined 1.3M popularity signal across the ranked entries.