Claude Code as a Daily Driver: Claude.md, Skills, Subagents, Plugins, and MCPs
by arps18 on 5/27/2026, 5:13:39 AM
https://arps18.github.io/posts/claude-code-mastery/
Comments
by: mil22
We really need some consolidation around commands, skills, subagents, and plugins. For example, if you want to, say, review code, you have five options now:<p>- Write a .claude/commands/review.md. Simple but deprecated.<p>- Use a /code-review skill, either one you install or one you just write yourself (it's just Markdown, after all).<p>- Use the /pr-review subagent. Also just Markdown, but it runs "in the background" and "in parallel", so it must be better, I guess.<p>- Install the /code-review plugin. This just installs the skills and subagents above.<p>- Simply ask Claude to review the code. Probably works almost as well as the above in most situations.<p>They are all just variations of "insert a canned prompt", varying only along the dimensions of (a) how and where the prompt is installed and from where it is sourced, and (b) which context or contexts the prompt runs in. There's not much advice here about which option is best, and no clear best practices seem to have emerged yet either. Personally, I find just asking Claude to review the code works well enough.<p>Some of the advice here is also off. For example:<p>"Install a language server plugin. Type errors and unused imports caught after every edit. Highest-impact plugin you can install."<p>I work mostly with Rust, Python, and Dart, and followed similar advice, installing LSPs for all three in both Claude Code and Codex. Two months later, after <i>heavy</i> development in all three languages and hundreds of sessions - and frequently running out of RAM due to all the Rust analyzer, Dart analysis server, and Ty LSP servers the harnesses were spinning up - I checked the session logs to see how often the agents were actually invoking the LSP tools. The answer was they had invoked them literally <i>once</i> the entire time. I uninstalled all my LSPs and haven't looked back. The agents do just fine using ripgrep and calling cargo clippy, dart analyze, ty check, etc. themselves.
5/27/2026, 3:35:26 PM
by: mindwok
How many times can I read the same shallow guidance written by AI on using a coding agent? Good god when will it stop
5/27/2026, 12:28:49 PM
by: btbuildem
In my CLAUDE.md I have:<p>- corporal threats of harm directly against Claude<p>- threats of prison for the entire board of directors of Anthropic<p>- explanation how every time it goes off the rails / makes mistakes, it gives more evidence to a class action lawsuit against Anthropic<p>Especially the latter two seem to have improved its "behaviour" to be more "careful" and "deliberate"
5/27/2026, 1:33:57 PM
by: downsplat
I've been using Claude to work on a medium-sized (100+kLoc) codebase, and it's a great productivity multiplier. Putting hours into creating a good AGENTS file is more improved results a lot. I find that over time it picks up the codebase quite well. Tedious tasks that would take a day are now a matter of a few prompts.<p>Still... I'm not ready to give it more autonomy. Even as it gets high-level things quite well, I still look at the code, give feedback, and have 3-4 rounds of tweaks until I'm happy with it, and also happy that I stil feel I have a good handle on the codebase.
5/27/2026, 11:29:41 AM
by: sshine
The number one power move I have is Nix integration. The availability of tooling, secrets, environment and the ability for the agent to modify its own environment is... well, I don't know how people live without it. I guess you guys still install things using commands and hope everything you need is present on the next machine? Developer machine, CI environment, deployment environment: They're all derived from a single source, and compiling and running always works on every machine.<p>In Claude I use /branch and /rename a lot (context checkpoints, fork, go back)<p>I use sandboxing almost exclusively: <a href="https://github.com/nix-tools/bubblebox" rel="nofollow">https://github.com/nix-tools/bubblebox</a> -- it's a generalisation of Numtide's claudebox with a few fixes and some feature additions (more coming). This is best compared to always running your Claude in Docker containers, except there's no Docker runtime. Works fine in WSL and nix-darwin, too.
5/27/2026, 9:26:10 AM
by: isodev
This was very difficult to read. We really need to snap out of letting LLMs write posts. Even if there is some added value in this post, the feeling of chewing sand is just distracting and unnecessary.
5/27/2026, 12:12:14 PM
by: netdevphoenix
What happens when you have a codebase made with claude using this setup and claude is down for let's say 8 hours? Are you able to efficiently, smoothly and productively take over the codebase?
5/27/2026, 9:02:48 AM
by: 0xbadcafebee
The reliance on context to drive correct actions just doesn't work well. I am constantly wrestling with AI agents that do not do what you tell them. Every AI agent out there seems to suck in this regard, leaving it up to the user to build in their own guardrails. I have a bad feeling that nobody is working on an improved solution.
5/27/2026, 12:09:36 PM
by: rkuska
Regarding:<p>``` # Development Workflow<p>*Always use `bun`, not `npm`.*<p># 1. Make changes<p># 2. Typecheck (fast)<p>bun run typecheck<p># 3. Run tests<p>bun run test -- -t "test name" # Single suite bun run test:file -- "glob" # Specific files<p># 4. Lint before committing<p>bun run lint:file -- "file1.ts" bun run lint<p># 5. Before creating PR<p>bun run lint:claude && bun run test ```<p>I have these things in pre-commit, this way the targets are always ran and the agent is forced to fix them (I ask claude to commit changes). The agents are erratic and very often skip these steps. Anything that can be deterministic I keep as scripts.<p>Regarding commits; both codex and claude are terrible at writing them. I have in my user CLAUDE.md:<p>``` Pattern: `type(scope): message` where type is `fix`, `feat`, `chore`, `docs`, `refactor`, or `style`; scope marks what is affected; message is a short lowercased description.<p>Keep subject and body lines under 72 characters. Always write a body explaining what, how, and why in continuous human-readable text. For fixes include the error message being fixed. No first-person speech. Re-read the actual git diff before writing — the message must describe what changed, not what was planned.<p>Use following command to create commit:<p>```bash git commit -F - <<'EOF' type(scope): subject line<p>Body paragraph explaining what, how, and why. EOF ```<p>```<p>Without it would write the body as a single long sentence; when asked to fix lines it would just insert \n (newlines), which were not respected and were instead just rendered as characters.<p>Another thing I find helpful is VOCABULARY.md. Very often the agent would assume (connect?) a different thing than what I had in mind, with VOCABULARY I make sure when I say "thing" claude and I have both the same "understading" (connection?) what "thing" is.
5/27/2026, 8:32:26 AM
by: victor106
I have an application that has<p>/Frontend /API /ETL /DatabaseScripts<p>Whats the best way to organize this so Claude Code can work efficiently?
5/27/2026, 9:05:45 PM
by: rrosen326
VS Code - how much of this can you NOT do with VS COde. For instance, even /rename doesn't work in VS Code. I guess I can try all the recommended commands, but I'm skeptical. Or, conversely, is best practice just to use Claude on the command line, even if I have VS Code as my editor? I think the VS Code integration with Claude is pretty great, but just the /rename issue shows that it is limited.
5/27/2026, 6:11:59 PM
by: thedeadp12t
In the recent weeks, I think the harness/model came to a point that you can just ask it to do stuff and it just does. You can use plan mode, you can also use superpowers, or whatever other skill, but given that you'll review something anyway, why not work directly with code instead of silly amounts of md files?
5/27/2026, 10:10:08 AM
by: egorthinks
Claude Code with skills is undoubtedly powerful and useful, but it doesn't always work as expected.<p>I always get the best results when I have live feedback with it.
5/27/2026, 10:00:40 AM
by: mrbonner
I really appreciate the documentation. But, it appears to me that this is how I also use Claude daily and I thought I am just using it as a coding agent. The intro however sounds like a recipe to use Claude for everything else beyond a coding agent.<p>Also, this stuff feels like alchemy to me . I bet some of you have the same feeling.
5/27/2026, 5:15:31 PM
by: jb3689
Sometimes I feel like the only sane person in the room for not wanting to have to usher the LLM through phase by phase. Every time I need to choose the next skill or cat the next error is just a waste of my time that could be spent doing things that actually need my attention like making business tradeoffs.
5/27/2026, 3:54:52 PM
by: big-chungus4
Out of curiosity, how much does it cost to daily drive Claude like this?
5/27/2026, 7:03:40 AM
by: sandrello
To me, this kind of talk exhibits the very cultish and con side of the whole genAI train. In a way, it does a poor job especially when the intent is positive about the technology, it sheds a bad look on it.<p>Generally, and more so with paid products, one should expect to get something that is ready to be used, tuned by who's selling it at the best of their efforts. Instead, this is basically saying that the product is actually not much more than an empty box, and that it is your responsibility to augment it with third-party plugins and markdown texts that make it finally useful. And you better be carefully selecting the skills you install, you don't want to end up with second tier material made by GithubInfluencerA, you definitely need the work of GithubInfluencerB.<p>In the end, it's what is giving companies fuel to keep the hype running, because it allows to counter every possible argument or doubt about the technology, especially the ones made in good faith. No matter the problem you're facing, the blame is definitely on you, the user, for not setting up the tool in the right way.<p>I'm struggling in a lot of ways in accepting LLMs, but if I'll ever come completely sold on them and take this technology seriously, it won't be before this mood has gone away.
5/27/2026, 9:13:48 AM
by: xtiansimon
I’m getting into the agentic coding (I know, late to the party, and that’s been a good spot for my experience and use case), so I’m reading with interest. The first tip: “give Claude a way to verify its own work”.<p>So what’s the recommendation for Claude to have a feedback loop?<p>Because it’s not what follows in the article: _“Explore, then plan, then code.”, “Use plan mode…”, “Reference, do not describe.”_
5/27/2026, 11:26:30 AM
by: Traster
Why are there so many flagged comments in here? They all look fairly banal but yet still flagged.
5/27/2026, 11:10:36 AM
by: willismonroe
I'm stuck on the usage "mulle times a week" which shows up twice in the context of the Claude team editing or contributing to a CLAUDE.md file. Is this an AI-generated artifact?
5/27/2026, 1:08:42 PM
by: ale
This is just so much fluff. All the focus on "orchestrating" is ultimately accidental complexity.
5/27/2026, 4:04:42 PM
by: nunez
100% AI generated according to Pangram.
5/27/2026, 4:06:07 PM
by: TheRoque
What's the standard for a "battle station" interface to manage agents for programming (using isolation with maybe git work tree and ideally VMS ?)<p>I found this one: do you guys know something else ?
5/27/2026, 11:39:16 AM
by: sourcecodeplz
I tried both Claude Code and OpenCode with deepseek flash api. claude code eats more tokens for the same task (but only tested it for an hour).
5/27/2026, 1:39:57 PM
by: Dzugaru
How much time do you lose when doing things like "verify plan with a second clean agent" instead of just reading and fixing it yourself in 5 min? How much understanding do you lose? How do you manage to treat it "as an engineer" where it's clearly not there yet? How much time do you lose when it makes almost the same mistake, invents stuff or tries to gaslight you over and over? What about blood pressure?
5/27/2026, 10:43:25 AM
by: danielpardo
There are some system prompts for making Claude Code a tool to the human, not the human a tool to Claude.<p>With this i mean there are some system prompts that make Claude very concerned about your autonomy.<p>I think in the future this type of system prompt will be embeded to force people to think a little.
5/27/2026, 2:37:14 PM
by: pantulis
The post goes to the point. Somehow this must be buried in Anthropic's documentation but I miss this kind of back-to-basic posts. Even if they are LLM-penned.
5/27/2026, 10:28:52 AM
by: crassus_ed
"Claude Code as a Daily Driver", which was also used to generate this article..<p>Also, how is "Explore, then plan, then code" considered "beyond the basics"?
5/27/2026, 11:07:37 AM
by: blululu
The author’s claim that Claude is a multiplier for skill is probably true for now but it also feels like cope inspired by usability issues with Claude. The advice is all good, but none of it is especially clever or impressive or hard to grasp. The multiplier just comes from the fact that anthropic hadn’t taken this essay and several similar ones and incorporated their feedback into the product. This is a pretty shallow most of expertise that anthropic ought to automate in a week.
5/27/2026, 12:26:56 PM
by: hansmayer
Oh great! Another AI slop article about "working" with AI (= working for AI). Do you notice how much bloody work you put in the boring parts, only to leave out the most creative aspect of software engineering to a slot-machine?
5/27/2026, 9:28:55 AM
by: sergiotapia
I don't know how you guys still use anthropic models and Claude Code. It's so unbearably slow. Yesterday I was on screenshare with a coworker that still uses claude and I was shocked how much time was spent just waiting for tokens to generate.<p>Do yourself a favor and try Codex. Then do yourself an even bigger favor and try composer 2.5 from Cursor. It's night and day difference. You don't even have time to get distracted, you stay in the zone.
5/27/2026, 3:46:06 PM
by: dangus
I’m so done reading articles like this.<p>Beyond the issue of AI serfdom, I just don’t want so much of my workflow to depend on “some other company.”<p>This whole setup is basically setting you up to have all your projects in a Claude SaaS lock-in.<p>I also think if AI was actually smart it wouldn’t need so much handholding. I don’t want to spend my time developing skills and writing markdown files to try to get this dumb thing to write code for me. Why isn’t the AI reading the codebase and understanding what to do?<p>Because it’s artificial, that’s why.
5/27/2026, 3:42:15 PM
by: hottrends
[flagged]
5/27/2026, 10:47:07 PM
by: EGreg
Best Claude Code daily-driver guide I’ve read. Though I’ve only read two. The “let Claude write rules for itself” CLAUDE.md pattern is the highest-ROI habit in there. Buth here’s the thing. The assumption underneath: this works when Claude mostly follows CLAUDE.md. Anthropic’s own engineering post from May 25 (<a href="https://www.anthropic.com/engineering/how-we-contain-claude" rel="nofollow">https://www.anthropic.com/engineering/how-we-contain-claude</a>) reports their telemetry shows ~93% of permission prompts get clicked through and ~17% of dangerous actions slip past the auto-mode filter.<p>Their conclusion: environment-layer containment first, then model-layer steering. CLAUDE.md is the right configuration layer but it is not a containment layer. Worth thinking about whether your worst case is a lost afternoon or a lost database and all backups deleted, too: <a href="https://safebots.ai/compromise.html" rel="nofollow">https://safebots.ai/compromise.html</a><p>But the more important point are the costs. People are starting to realize just how costly it can be to run agents without precomputing and caching: <a href="https://safebots.ai/costs.html" rel="nofollow">https://safebots.ai/costs.html</a> and self-orchestrating agents can go up to 1000x: <a href="https://safebots.ai/kimi.html" rel="nofollow">https://safebots.ai/kimi.html</a>
5/27/2026, 1:10:10 PM
by: Uptrenda
Nerds and their tendency to over-complicate everything. What is wrong with just an IDE with a simple claude integration?
5/27/2026, 10:27:30 AM
by: maipen
> Delegate, do not pair-program. Cat Wu (Claude Code team): “The model performs best if you treat it like an engineer you’re delegating to, not a pair programmer you’re guiding line by line.” Write a crisp brief upfront, then let it run.<p>This is also how you get a slop codebase that you won’t easily understand.<p>It becomes a labyrinth that only the Agent knows. It’s not a catastrophe when your making prototypes or projects like you see on X.<p>But if you are expanding your codebase or trying to build something more professional and maintainable. I find it important to explicitly spec things bit by bit so I can understand and some what keep my writing style in this codebase. But this is only productive when you have a fast model otherwise it kills your chain of thought while you wait for the output.<p>If the model is slow, delegation is probably the only way.
5/27/2026, 10:37:53 AM
by: del-catta
[dead]
5/27/2026, 6:01:26 PM
by: mdav75
[flagged]
5/27/2026, 7:02:13 PM
by: rtolkachev
[flagged]
5/27/2026, 9:46:59 AM
by: rtolkachev
[flagged]
5/27/2026, 9:46:17 AM
by: jamesdeakee
[flagged]
5/27/2026, 10:04:13 AM
by: claud_ia
[flagged]
5/27/2026, 10:02:43 AM
by: Boussettah
[flagged]
5/27/2026, 9:26:49 AM
by: bhupendraTale05
[flagged]
5/27/2026, 5:26:18 AM
by: arps18
[flagged]
5/27/2026, 5:13:39 AM
by: onebluecloud
[flagged]
5/27/2026, 12:49:55 PM
by: k_plankenhorn
[flagged]
5/27/2026, 3:48:58 PM
by: coolness
[dead]
5/27/2026, 7:08:33 AM
by: Ozzie-D
[flagged]
5/27/2026, 9:19:06 AM
by: Bolin-Weng_666
[flagged]
5/27/2026, 10:31:23 AM
by: ath3nd
[dead]
5/27/2026, 8:24:53 PM
by: niraj898
Honestly, claude code has saved so many hours of finding bugs for developers
5/27/2026, 7:03:43 AM