Skip to content
ReviewRamble

The weekly brief

AI, reviewed, in your inbox.

  • New reviews of tools, models, MCP servers and repos, with the verdict up front
  • Re-checked prices and ratings when something changes
  • New analysis pieces as they are published

Models, agents, MCP servers and the tools actually worth your time.

One email a week. Unsubscribe with the link in any brief. How we use your email

The best Claude Code plugins and agent skills in 2026

The best Claude Code plugins and agent skills right now are Superpowers, Matt Pocock's Skills and Anthropic Skills, out of 55 we have reviewed. Each is ranked by our rating, with its price, our verdict and who it suits.

By ReviewRamble editors · Updated Nov 18, 2026 · How we review

How we ranked: by our rating out of 5, from reviews built on each product's own pricing, documentation and policies plus independent benchmarks where they exist, each checked on the date shown in the review. We take no affiliate commission.

  1. 1

    A discipline layer that makes a coding agent behave like a careful junior engineer, and worth its extra tokens on anything bigger than a one-line fix.

    Best for: Developers who want an agent to run multi-hour features with a plan, tests and review built in.

    Price: Free and open source (MIT). You pay for the model usage of the agent you run it in.

    Read our Superpowers review
    4.7/5
  2. 2

    Matt Pocock's skills make the agent argue the plan out with you before it writes code, and /grill-me alone justifies the install.

    Best for: Engineers who want lightweight planning and TDD habits without handing the process to a framework.

    Price: Free and open source (MIT). You pay for your agent's model usage.

    Read our Matt Pocock's Skills review
    4.5/5
  3. 3

    Read it before writing a skill of your own, and install it if Claude Code has to handle Office files and PDFs; just don't mistake the document skills for something you can fork.

    Best for: Anyone writing skills, and Claude Code users who need Office and PDF handling.

    Price: Free. Example skills are Apache-2.0; the document skills are source-available, not open source.

    Read our Anthropic Skills review
    4.4/5
  4. 4

    A generic "check for OWASP issues" prompt is a guess; this is how a working audit firm actually looks for bugs, written down for your agent.

    Best for: Teams that want structured, credible security review from their coding agent before risky merges.

    Price: Free under CC BY-SA 4.0.

    Read our Trail of Bits Skills review
    4.4/5
  5. 5

    Addy Osmani's pack turns senior-engineer habits into gates an agent cannot talk its way past, and it is the rare skill set that takes the browser seriously.

    Best for: Web developers who want one structured skill pack from spec to launch.

    Price: Free and open source (MIT). You pay for your agent's model usage.

    Read our Agent Skills (Addy Osmani) review
    4.3/5
  6. 6

    Start here before adding any third-party marketplace: commit-pinned entries and visible context cost make it the safest default, though a listing is not a security audit.

    Best for: Every Claude Code user, as the first place to look for plugins.

    Price: Free. The catalogue repo is Apache-2.0; each plugin has its own licence, and many vendor plugins need a paid account with that vendor.

    Read our Claude Code official plugin marketplace review
    4.3/5
  7. 7

    Diagram Design gives agents the taste they lack, and its diagrams look published rather than generated, at the price of an output you cannot diff or drag.

    Best for: Writers, founders and engineers who publish diagrams in posts, decks and docs and want them to match their brand.

    Price: Free and open source (MIT). You pay only for your agent's model usage.

    Read our Diagram Design review
    4.3/5
  8. 8

    A real detector and a shared design vocabulary make this the design skill worth installing, provided you switch off telemetry and keep it the only design skill in the project.

    Best for: Developers whose agents build frontend UI and want it to look less generic.

    Price: Free and open source (Apache-2.0).

    Read our Impeccable review
    4.3/5
  9. 9

    Spec Kit turns agent work into reviewable documents in git, and charges for it in hours of human reading per feature.

    Best for: Teams that want written, reviewable specs before an agent writes code.

    Price: Free and open source (MIT).

    Read our Spec Kit review
    4.3/5
  10. 10

    EGM's zero-context reviewers are the cleverest answer yet to an agent rubber-stamping its own design, as long as your harness can spawn the subagents it needs.

    Best for: Teams using Antigravity, Claude Code or Codex that want a design reviewed by something other than the agent that wrote it.

    Price: Free and open source (Apache 2.0).

    Read our Elephant-Goldfish Model review
    4.2/5
  11. 11

    If your agent touches Google Cloud, these are the instructions it should be reading, but read the commands they tell it to run before you let it run them unattended.

    Best for: Engineers who let coding agents work in Google Cloud projects and want Google's own procedures loaded instead of the model's memory of gcloud.

    Price: Free under Apache-2.0. The Google Cloud, Ads and Analytics services the skills drive are billed as usual.

    Read our Google Agent Skills review
    4.2/5
  12. 12

    For a Claude Code user this is a one-command security win, with a reviewer that is not grading its own homework.

    Best for: Claude Code users who want security review on every change without setting anything up.

    Price: Free plugin under Apache 2.0; the two model-backed reviews use your Claude or API usage.

    Read our security-guidance plugin (Claude Code) review
    4.2/5
  13. 13

    SwiftUI Pro is the cheapest fix for an agent that keeps writing outdated SwiftUI, provided you accept rules that age with every Apple release.

    Best for: iOS and macOS developers who let an agent write SwiftUI and want its output to match current Apple APIs and accessibility rules.

    Price: Free and open source (MIT). You pay only for your agent's model usage.

    Read our SwiftUI Pro review
    4.2/5
  14. 14

    Worth installing on large repos where your agent keeps grepping, provided you treat the benchmark numbers as the vendor's own and choose the docs backend deliberately.

    Best for: Teams with large codebases who want their agent to navigate structure instead of grepping.

    Price: Open-source engine free (Apache-2.0, with an MIT licence file also in the repo). Graphify Cloud has a free hosted tier and per-seat Pro and Teams plans with prices not published; self-hosted Enterprise is in early access.

    Read our Graphify review
    4.1/5
  15. 15

    Hallmark turns design taste into rules an agent can fail, which makes its pages look made rather than generated, at a context bill nobody has yet measured.

    Best for: Developers who build marketing pages and landing pages with Claude Code, Codex, Cursor or OpenCode and want them to stop looking like every other generated site.

    Price: Free and open source (MIT). An optional TOGETHER_API_KEY enables generated images, billed by Together AI.

    Read our Hallmark review
    4.1/5
  16. 16

    Playwright's own test agents give any coding agent a disciplined plan, write, heal loop that ends in ordinary Playwright code you own, with no vendor in between.

    Best for: Teams already using Playwright who want a coding agent to draft and maintain end-to-end tests without adding a vendor.

    Price: Free and open source (Apache-2.0). You pay for the coding agent and model that run the agents.

    Read our Playwright Test Agents review
    4.1/5
  17. 17

    A well-argued brake on agents that over-build, with honest guard rails, though its headline savings come from the author's own benchmark.

    Best for: Developers whose coding agent writes more code and adds more dependencies than the task needs.

    Price: Free and open source (MIT). No paid tier; you pay for your agent's model usage.

    Read our Ponytail review
    4.1/5
  18. 18

    Scientific Agent Skills gives a general coding agent the conventions a working scientist would insist on, provided you install the few skills you need rather than all 177.

    Best for: Researchers and lab engineers who already use a coding agent and want it to follow field conventions for specific packages and databases.

    Price: Free and open source (MIT). You pay for your agent's model usage and any licensed databases. K-Dense's hosted K-Dense Web is separate: Plus $199/month, Team $499/month, pay-as-you-go runs up to $9 (Standard) or $29 (Pro).

    Read our Scientific Agent Skills review
    4.1/5
  19. 19

    A skill that turns a concept into a machine readers will stop and play with, and charges for it in long, token-hungry sessions.

    Best for: Technical writers and developers who want interactive explanatory figures for docs, posts and landing pages.

    Price: Free and open source under MIT. Runs on the agent subscription or API you already use.

    Read our Anatomy review
    4.0/5
  20. 20

    Diagrams that check themselves before they reach you, which is exactly what agent-drawn architecture pictures have lacked.

    Best for: Engineers and tech leads who want their coding agent to produce architecture and flow diagrams that are checked, interactive and shareable as one file.

    Price: Free and open source (Apache-2.0; earlier releases MIT).

    Read our Archify review
    4.0/5
  21. 21

    A sound way to put a second model's review on Claude's work, held back by a maintainer queue that has not moved since July.

    Best for: Claude Code users with a ChatGPT or OpenAI account who want a second-model review.

    Price: Free and open source (Apache-2.0). Usage counts against your ChatGPT plan or OpenAI API billing.

    Read our Codex plugin for Claude Code review
    4.0/5
  22. 22

    A genuinely useful map for people who built an app with AI and cannot read it, though at one week old it is still a young plugin from one author.

    Best for: Founders and small teams who built an app with Claude Code or Codex and need to understand, explain or sanity-check it without reading every file.

    Price: Free under the MIT licence. Map generation uses your own Claude or ChatGPT plan; no separate API key or fee.

    Read our Floorplan review
    4.0/5
  23. 23

    A thoughtful, well-guarded setup layer for people who run several coding agents, held back by release churn and documentation few will read to the end.

    Best for: Developers or small teams running two or more coding agents who want shared project memory and one review standard across them.

    Price: Free and open source (MIT). You pay for the model usage of the agents it configures.

    Read our Gentle-AI review
    4.0/5
  24. 24

    Humanizer is a sharp, honest editing skill that strips the obvious AI tells from your prose, as long as you remember it is editing for readers, not for detectors.

    Best for: Writers and developers who want obvious AI tells removed from drafts and documentation while keeping the meaning intact.

    Price: Free and open source (MIT). You pay only the model tokens your agent spends running it.

    Read our Humanizer review
    4.0/5
  25. 25

    text-to-cad turns a coding agent into a capable junior CAD engineer by making it write real parametric code, but it changes fast and reports usage by default.

    Best for: Engineers and makers who can read Python and want an agent to draft printable or machinable parts.

    Price: Free and open source (MIT). You pay for your agent's model usage and any fabrication service you send parts to.

    Read our text-to-cad review
    4.0/5
  26. 26

    tstack packages a working engineer's pull-request routine into skills you can copy, and its multi-model review loop is the useful idea, though full-send merges code nobody watched.

    Best for: Developers who already let agents open pull requests and want a repeatable routine for reviewing, fixing and merging them.

    Price: Free and open source (MIT). You pay for the agents and models it calls.

    Read our tstack review
    4.0/5
  27. 27

    A free catalogue of specialist subagents that rewards picking two plugins and reading them, and punishes installing all 94.

    Best for: Developers who want ready-made specialist subagents for their language or stack.

    Price: Free and open source (MIT). You pay for your agent's model usage.

    Read our wshobson/agents review
    4.0/5
  28. 28

    The rare AI audit tool that assumes its own findings are wrong until another agent fails to knock them down.

    Best for: Security teams auditing a codebase who want verified findings rather than a long list of maybes.

    Price: Free under MIT; audits use a lot of model tokens.

    Read our Cloudflare Security Audit Skill review
    3.9/5
  29. 29

    The safest way we have seen to let a coding agent loose on footage, wrapped in more tools and more releases than most people will ever need.

    Best for: Developers and small video teams who already use Claude Code, Cursor or Codex and want the agent to edit recordings without hand-written FFmpeg commands.

    Price: Free and open source (MIT). Needs FFmpeg 6.1+ and Python 3.10+; no API keys or cloud service.

    Read our ffmpeg-skill review
    3.9/5
  30. 30

    A diagram skill that checks its own geometry before handing over, ideal for slides and posts, though Mermaid still wins for diagrams that live in the docs.

    Best for: Engineers and technical writers who need polished architecture diagrams for slides, posts and design reviews from inside Claude Code or Codex.

    Price: Free and open source (MIT). Runs on the agent subscription or API you already use.

    Read our fireworks-tech-graph review
    3.9/5
  31. 31

    Deep research with real verification gates instead of a polished summary, paid for in hours of agent time and a pipeline you have to trust one maintainer to keep working.

    Best for: Analysts and researchers already on Claude Code who need cited reports they can audit and are willing to wait hours for them.

    Price: Free and open source (MIT); you pay for the Claude Code or Codex usage it consumes. Hosted version: pay as you go, $9 per Light report and $49 per Deep report; Enterprise on request.

    Read our hyperresearch review
    3.9/5
  32. 32

    A sensible always-on rule set you can read in two minutes; useful, but it is a starting file, not a tool.

    Best for: Developers who want a short, ready-made CLAUDE.md that discourages over-building.

    Price: Free. The README says MIT, but the repo has no LICENSE file and GitHub detects no licence.

    Read our Karpathy-inspired CLAUDE.md review
    3.9/5
  33. 33

    universal-modder turns a coding agent into a capable modding assistant for single-player games, and its shared field notes matter more than the art pipeline.

    Best for: PC players and hobby developers who want an agent to build mods for single-player games they own.

    Price: Free and open source (MIT). You pay for your coding agent and for fal generations (per-generation pricing; um fal price shows the cost). Local ComfyUI is free; Sprixen and Voidless are separate paid services.

    Read our universal-modder review
    3.9/5
  34. 34

    An awesome list that earns the name by telling you what each mod can touch, which matters more than the count.

    Best for: Claude Code users who want to find a mod for a specific job and check what it can access before reading its source.

    Price: Free (CC0). The catalogue, website and mod-finder plugin cost nothing.

    Read our Awesome Claude Code Mods review
    3.8/5
  35. 35

    BMAD earns its paperwork on projects with stakeholders who sign things off, and charges for it in hours on everything smaller.

    Best for: Teams on larger projects that want PRDs, architecture and stories generated and reviewed.

    Price: Free and open source (MIT). "BMad" and "BMAD-METHOD" are trademarks of BMad Code, LLC.

    Read our BMAD Method review
    3.8/5
  36. 36

    claude-obsidian treats a notes vault like a database, with transactions and citations, which makes it trustworthy and slow to adopt.

    Best for: Obsidian users who want Claude Code to build and query a cited research wiki.

    Price: Free and open source (MIT). You pay for your agent's model usage.

    Read our claude-obsidian review
    3.8/5
  37. 37

    A serious, well-ruled playbook for agent-built co-op mods, whose shared directory and test tools are worth more than any single game it has finished.

    Best for: Patient PC players and modders with a coding agent who want to revive a stalled co-op mod or attempt co-op for a game nobody has covered.

    Price: Free and open source (MIT). The cost is your coding agent's usage; every player needs their own copy of the game.

    Read our Co-op Anything review
    3.8/5
  38. 38

    Diorama turns a vague request for a pretty 3D scene into a disciplined build with research, a fixed look and frame-by-frame checks, and asks a lot of the agent's context in return.

    Best for: Developers who want a coding agent to produce animated Three.js scenes in a consistent pixel or toon style and prove each frame works.

    Price: Free and open source (MIT). You pay for the coding agent and model that use it.

    Read our Diorama review
    3.8/5
  39. 39

    ECC is a parts bin rather than a workflow: useful in a minimal profile, a standing tax on context at full size.

    Best for: Claude Code power users who want a large toolkit and will trim it to a minimal profile.

    Price: Free and open source (MIT). ECC Pro GitHub App for private repos is $19 per seat a month; Enterprise is quoted. You pay for your agent's model usage.

    Read our Everything Claude Code (ECC) review
    3.8/5
  40. 40

    last30days does the half-hour of Reddit and X trawling you would otherwise do before a meeting, as long as you are comfortable pointing a scraper at platforms through your own logins.

    Best for: Researchers, founders and marketers who want a fast, engagement-ranked read on a person or topic and will wire up official platform APIs.

    Price: Free and open source (MIT). You supply, and pay for, any provider keys and reranking model you turn on.

    Read our last30days review
    3.8/5
  41. 41

    reanimate fixes the real reason agents botch animation clones (they cannot see motion) by making the agent measure its copy against the original until the numbers agree.

    Best for: Front-end developers using Claude Code who need a specific hover, entrance or WebGL effect reproduced closely and want proof of how close it is.

    Price: Free (MIT). No API keys; the cost is the Claude Code usage of runs that often last 30 to 90 minutes.

    Read our reanimate review
    3.8/5
  42. 42

    A sharper idea than most spec tools, a persistent product model rather than throwaway specs, carried by a 97-star project one person maintains.

    Best for: Teams on coding agents who want one living product definition in git rather than a spec written and discarded per feature.

    Price: Free and open source (MIT).

    Read our pdd-kit review
    3.7/5
  43. 43

    Bang Motion gives an agent a menu of looks and structures that keeps its videos from resembling slide decks, but its rules are a designer's opinions rather than measurements, and the repo stopped moving in mid-September.

    Best for: Developers and solo marketers who want an agent to produce varied web-native openers and explainers with minimal setup, and who can judge the motion by eye.

    Price: Free under the MIT licence. Costs are agent tokens, plus any image or video MCP you choose to call for assets.

    Read our Bang Motion review
    3.6/5
  44. 44

    A carefully maintained studio costume for Claude Code that makes a solo developer write the documents a team would, which helps on a long project and burns quota on a short one.

    Best for: Solo developers or tiny teams on a months-long Godot, Unity or Unreal project who want design docs, stories and QA evidence kept in order and can afford a Max-level Claude plan.

    Price: Free and open source (MIT). All cost is Claude Code usage on your Claude plan or API key; three director agents request Opus.

    Read our Claude Code Game Studios review
    3.6/5
  45. 45

    A deploy skill with unusually careful guard rails, still an alpha with one author and only a handful of providers tested live.

    Best for: Solo builders whose coding agent made an app on a common stack and who want it live on their own Vercel or Netlify, Supabase or Neon and domain accounts.

    Price: Free and open source (MIT). Hosting, database, domain and email costs are billed by your own providers.

    Read our GoLive review
    3.6/5
  46. 46

    A rare animation skill with real engineering under it, the gait solver above all, but four days old and built for one look.

    Best for: Front-end developers and designers using Claude Code who want interactive, hand-drawn-looking SVG figures or scenes on a web page.

    Price: Free under MIT. Using the skill costs whatever your Claude plan costs; rendering video needs a local Chrome and FFmpeg.

    Read our line-art-animation review
    3.6/5
  47. 47

    motionmaxxing writes down more real motion-design craft than any video skill we have read, then leaves it as one unmaintained upload checked by a single rater.

    Best for: Developers who already make product films with an agent in HTML and GSAP and want the output to move like a motion designer's, and who can live with a repo that may not be maintained.

    Price: Free and open source (Apache-2.0; bundled GSAP under its own licence). An optional ElevenLabs key enables voice, sound effects and music, billed by ElevenLabs.

    Read our motionmaxxing review
    3.6/5
  48. 48

    A carefully written LinkedIn ghostwriting kit that knows its limits better than most, held back by one creator's house style and scraping routes LinkedIn's own terms forbid.

    Best for: Founders, consultants and marketers who write their own LinkedIn posts with Claude or Codex and want drafts in their voice, not a publishing tool.

    Price: Free and open source (MIT). Optional routes use your own Apify account (about $0.005 per LinkedIn post fetched) and Gemini API key.

    Read our social-media-skills review
    3.6/5
  49. 49

    OpenSpace asks the right question about agent skills, which ones actually work, but it lets agents rewrite their own skills by default and has gone quiet since August.

    Best for: Teams running agents against a large, growing skill library who want evidence about which skills work and will set evolution to audit-only first.

    Price: Free and open source (MIT). The open-space.cloud hub is optional and no paid plan is published. You pay your model provider.

    Read our OpenSpace review
    3.5/5
  50. 50

    Three excellent example machines and a sound method for building more, though what Claude produces from a one-line request is likely to fall short of the demos.

    Best for: Engineers, teachers and technical writers who use Claude Code and want interactive cutaway explainers of real mechanisms, and are willing to iterate past the first draft.

    Price: Free under MIT. Building a machine uses your Claude Code plan or API usage; rendering runs locally.

    Read our teardown review
    3.5/5
  51. 51

    The Agency is the biggest shelf of agent personas on GitHub, useful as a starting library and weaker than its 168,000 stars suggest once you read the files.

    Best for: Developers and small teams who want ready-made subagent prompts for unfamiliar specialties and will edit the few they keep.

    Price: Free under MIT, as is the desktop installer app. You pay only for the agent or model plan you run the prompts on.

    Read our The Agency (agency-agents) review
    3.5/5
  52. 52

    Understand Anything gives new joiners a genuinely useful map of a mid-sized codebase, if you can afford the tokens for the first pass.

    Best for: Teams onboarding people to a codebase of a few hundred to a few thousand files who already pay for an agent subscription.

    Price: Free and open source (MIT). Analysis runs on your agent's model, so cost is your subscription quota or API tokens.

    Read our Understand Anything review
    3.5/5
  53. 53

    A catalogue worth searching and not worth installing whole: the review process is real, but the licence bookkeeping for its imported skills has not caught up.

    Best for: Developers and platform teams who want to search a large pool of agent skills, read the candidates and then install a handful, ideally from their upstream source.

    Price: Free and open source (MIT for code, CC BY 4.0 for original prose; imported skills keep upstream licences). You pay for your agent's model usage.

    Read our Agentic Awesome Skills review
    3.4/5
  54. 54

    Boss Skill is BMAD on autopilot: thorough paperwork and real test gates, paid for with a long, token-hungry run that you have to remember to stop.

    Best for: Developers who want BMAD-style PRDs, architecture notes and QA records produced automatically on new features, and can afford long agent runs.

    Price: Free and open source (MIT). Model usage is billed by your coding agent's provider.

    Read our Boss Skill review
    3.4/5
  55. 55

    A careful, well-tested day-trading toolkit whose own Time Machine calls its headline trade a coin flip, which is the most useful thing it will tell you.

    Best for: Retail traders who use Claude and want a structured, paper-first way to test intraday ideas against history before risking any money.

    Price: Free under MIT. Needs a Claude plan for the skills; data sources are free. Trading costs are your broker's: Alpaca paper trading is free, and live trading puts your own money at risk.

    Read our Day Trader skill review
    3.4/5

Sources

The main official source behind each ranked review; every review lists its full set.

More guides