Welcome back to the AI Edge weekly newsletter!
This week, we’re covering one of the most useful (yet underrated) Claude features that most people still don’t know exist.
Late last year, Anthropic shipped Claude Code Plugins.
With Plugins, you can install a full specialist agent into your Claude workspace in under 60 seconds.
In this guide, we went through BuildWithClaude and picked five of the most useful agents you can install right now.
By the end of this read, you’ll know exactly how to install them and what each one is best used for.
Special announcement
Our exclusive AI Edge Skool community is live now!
A 100% free online community with AI resources, guides, exclusive posts from Miles, and more.
Join here, and you’ll be one of the first members in the group - lots of free value!
With that out of the way, let’s get right into things!
Table of Contents
How to Install These Agents in 60 Seconds
All five agents in this guide were sourced from BuildWithClaude.
It’s a free public resource for finding Claude Skills, Plugins, agent files & more - we recommend you bookmark it.
The easiest way to install our list of agents below is through Claude Cowork.
First, add the BuildWithClaude marketplace:
/plugin marketplace add davepoon/buildwithclaude
You only need to do this once.
Claude Cowork → Customize → Plugins → Add
Then install the agent you want using:
/plugin install [plugin-name]@buildwithclaude
That’s it. Once installed, Claude can call the agents automatically when needed, or you can ask for one directly.
These agents come with their own skills, files, and connectors - think of each one as a mini AI employee already trained for a specific task.
Before installing anything, always verify that the Plugin is secure and check what it can access (especially if you’re going to download more agents from their site).
#5: Design Experience Agent
If you use Claude to build websites, apps, or interfaces, this is a useful one to have.
Design Experience gives you three specialist agents in one install:
→ Design Experience
→ Accessibility Specialist
→ UI/UX Designer
Instead of having one general agent handle the entire design process, you can use specialists for the interface, user experience, and accessibility.
Install:
/plugin install agents-design-experience
It is a simple way to improve your builds without having to do all design work yourself.
#4: Media Content Agent
This is the deepest bundle on the list.
Media Content comes with 12 specialist agents for anyone running a YouTube channel, podcast, or full content pipeline.
It covers things like:
→ Content production
→ Transcription
→ Trend research
→ SEO
+ more.
If you create content regularly, this can turn one Claude setup into a full content team.
Highly recommend.
Install:
/plugin install agents-media-content
#3: Sales & Marketing Agent
Probably the most immediately useful bundle here if you run a business.
This Plugin agent comes with eight specialists covering a full funnel:
→ Content marketing
→ Copywriting
→ Customer support
→ Sales automation
→ Social media content
→ X/Twitter management
Instead of using a single general Claude chat for everything, you can outsource each part of your sales/marketing workflow.
Install:
/plugin install agents-sales-marketing
#2 Codex Orchestrator
This one is very interesting.
Codex Orchestrator gives Claude direct access to OpenAI’s Codex.
Essentially:
Claude = project lead
Codex = dispatched workhorse
Claude can plan the task, send coding work to Codex, and review the output through checkpoints before moving forward.
If you already use both Claude Code and Codex, this brings them into one central hub.
Install:
/plugin install codex-orchestrator
#1 Claude GitHub Tools
Every vibe-coder & developer needs this.
Claude GitHub Tools brings a set of GitHub workflows directly into Claude Code.
It can handle:
→ Issue creation
→ PR review triage
→ Backlog summaries with quick-win suggestions
→ Release status checks
→ GitHub Actions authoring
You’ll need GitHub CLI installed and connected to your account before using it.
Instead of constantly moving between GitHub and Claude, you can handle much more of the workflow directly inside Claude Code.
That’s it!
Final Tips
A few quick things to round things out.
We recommend starting with the one agent closest to something you already do repeatedly.
In Claude Cowork, you can create your own agent Plugins (a very underrated feature that most people overlook)
Feel free to customize the listed agents to best fit your needs (add skills, files, etc.).
BuildWithClaude is 100% free and also has custom Claude skills that anyone can get started with - feel free to explore the entire site!
Midweek Edge
Our manually curated list of the most important news updates across AI, robotics & tech.
Cursor Launches Origin (GitHub Alternative)
Cursor just launched Origin, its own code hosting platform built directly into Cursor.
Think of Origin like an all-in-one GitHub alternative.
SpaceX Acquires Cursor
Cursor is now officially part of SpaceX ($60B acquisition).
SpaceXAI Launches Grok Bot
SpaceXAI launched Grok Bot, and it's one of the best agent tools on the market. A team of AI agents that can complete real work for you across apps and websites.
Available now with a SuperGrok heavy subscription.
Claude Code /design
Claude Code can now design your UI before it starts building it.
Run /design, and Claude will generate editable artboards with different design options.
Nous Research Releases Hermes Bot Mode
Nous Research recently released Bot Mode for Hermes.
With Bot Mode, you can set up several AI bots equipped with distinct roles, individual memory, custom routines, and dedicated chats.
Once configured, these bots can collaborate, assign tasks to one another, and coordinate their work.
Open Hermes Desktop → Bots and create your first AI team.
Google Releases Gemini 3.7 Flash
Google just dropped Gemini 3.7 Flash, its new model for coding, agents, and everyday knowledge work.
It improves on 3.6 Flash across coding and longer workflows, while launching at half of 3.6 Flash’s original price:
→ $0.75 per million input tokens
→ $3.75 per million output tokens
Try it now in Google AI Studio, Antigravity, or Gemini Spark.
GPT-5.6 Sol Ultrafast Mode
OpenAI introduced Ultrafast mode for GPT-5.6 Sol.
It runs Sol up to 14× faster than standard processing, reaching up to 750 output tokens per second without switching to a smaller model.
It is currently available to selected API users, with wider access coming later.
Alibaba Releases Qwen3.8-27B
Alibaba recently released Qwen3.8-27B, a new open-weight multimodal model small enough to run on consumer hardware.
→ 27B parameters
→ Understands text, images, and video
→ 262K native context, expandable to 1M
→ Can run locally on a laptop when quantized
Download it through Hugging Face or try it through Qwen.
Looking Ahead
Our manually curated list of the top AI trends, research, workflows & more.
Anthropic Model 2?
Anthropic quietly revealed an unreleased internal model, “Model 2,” in its latest risk report.
The company says it is slightly more capable overall than Mythos 5, although performance varies by task.
It is currently used internally and has not been released publicly.
AI Cost-Efficiency Hack
Grok 4.6 currently costs just $2 per million input tokens and $6 per million output tokens.
Based on current performance comparisons, that puts Fable-class and GPT-5.6 Sol-class intelligence at a fraction of the price.
We highly recommend using Grok 4.6 as your daily driver model, as it’s significantly cheaper than Fable/GPT-5.6.
Grok 4.7 Incoming
Elon Musk says Grok 4.7 should arrive within the next three to four weeks.
He says the model is already significantly better than Grok 4.6 and expects it to outperform current frontier models.
There are no independent benchmarks yet, so this remains a watch item until the model actually launches.
How to Be Successful (By Sam Altman)
Save this one.
Closing out
If you made it this far, thank you for reading, and we hope you found this week’s edition valuable.
If you enjoy reading, please forward our newsletter to someone you think would benefit from it.💙
Our promise to you: Every Wednesday, at 7 am EST, we’ll cut through the AI noise and send you human-curated AI content to make sure you stay ahead.
Content Pipeline
YouTube
Live Now: Grok Bot (full guide)
X (Twitter)
Live Now: How to Connect Hermes Agent + Buzz (step-by-step guide)
Free Skool Community
As a reminder, we just launched the free AI Edge Skool community, where we go behind the scenes and share more free value. Regular livestreams from Miles, free AI guides, a resource library & more.
If you want more, come join us here.
See you next Wednesday!
Interested in sponsorship or a partnership? Get in touch at [email protected]