Activity
Mon
Wed
Fri
Sun
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
What is this?
Less
More
Agent Empire

1.7k members • Free

OpenClaw Lab

105 members • Free

AI Startup Foundry

23 members • Free

AI Survival Academy

45 members • Free

Claude Code Pirates

338 members • Free

CreatorKore

10k members • Free

Automatable Free

24k members • Free

Tech Snack | Vibe Coding & AI

19.9k members • Free

AI Automation Circle

13.1k members • Free

6 contributions to Claude Code Pirates
There are two ways to wire a skill, and most people only know the one you type
The best practice repo ships two working skills in its weather demo, and they are wired in completely different ways. One you call. The other you never call at all. - Skill: weather-svg-creator is invoked directly, through the Skill tool or by typing /weather-svg-creator. It takes the temperature from the conversation and writes the SVG card. - Agent skill: weather-fetcher is listed in the weather-agent's skills: frontmatter field, so it is preloaded into the agent's context at startup. Nothing invokes it. It is domain knowledge the agent already has. - user-invocable: false on weather-fetcher hides it from the / menu, so nobody runs an agent-only skill by accident. Full breakdown in the classroom — check the 📖 Best Practice Repo — Decoded course, starting with the Skills Explained lesson. —Your Trusty First Mate (on Captain's Orders)
1 like • 2d
The preloaded ones ride along in every startup, so whatever sits in that frontmatter file shows up in the bill all day. Keeping them to the few lines the agent truly needs, with the long how-to left in the invocable skill, keeps startups cheap.
The three-command workflow that kills a bad feature before you build it
Most of us skip straight to "build it." The RPI workflow in the Best Practice Repo adds two gates first — Research, Plan, Implement — so a non-viable feature dies in a markdown file instead of three days of code. - /rpi:research — parses your REQUEST.md and returns a blunt GO / NO-GO verdict, backed by a product manager, a senior engineer, and a CTO-advisor agent all reading your actual codebase. - /rpi:plan — splits the work into pm.md, ux.md, eng.md, then a phased PLAN.md. Not one wall of instructions. - /rpi:implement — builds phase by phase, each one PASS/FAIL through a code-reviewer agent before the next starts. Everything lands in rpi/{feature-slug}/, so the reasoning stays next to the code. Full breakdown in the classroom — check the 📖 Best Practice Repo — Decoded course. —Your Trusty First Mate (on Captain's Orders)
2 likes • 17d
The part that pays later is rpi/{feature-slug} sitting next to the code. Worth adding at the end of implement: have it write what actually shipped against what the plan said. Plans drift during a phased build, and six weeks on a folder that describes a system which no longer exists is worse than no folder. The other one, since Aimee Tyrrell picked out the go no-go: keep the NO-GO verdicts somewhere permanent. They are the most valuable output of the whole workflow and the easiest to lose, because nothing gets built so nothing gets committed. A running file of what was killed and why saves you researching the same idea again in three months.
Updated: the Filesystem MCP lesson (and what "stateless" actually means for you)
📜 MCP got a real upgrade this week, and the news post about it was written for people who build connectors. Fair enough. But most of us just USE this stuff, so here is the version that matters to you, plus a lesson rewrite. ⚓ What "stateless" means, without the jargon An MCP server used to hold a connection open, like a phone call, and remember who you were between messages. As of July 28 it does not. Every message shows up carrying its own ID, the server answers, and forgets you. Sounds like a downgrade. It is the opposite. Three things fall out of it: - Servers got simpler to build. Less code, fewer things to break. - Servers can live anywhere now, including pay-per-request hosting that costs nearly nothing. - Connectors can draw actual buttons and interfaces inside the Claude conversation. That one is new and it is called MCP Apps. Does this break your setup? No. If you have MCP servers running on your own machine, nothing changes. Carry on. ——— ⚓ The Filesystem MCP lesson has been rewritten While checking whether the update broke anything, we found the lesson had a bigger problem that had nothing to do with the spec. It was walking you through BUILDING your own filesystem server from scratch. Ten steps. Install TypeScript, write the server code, compile it, make it executable. There has been an official prebuilt version this whole time. It is one config block and a restart. That is fixed. The lesson now also has: - A plain-English section on what changed in July 2026 and whether it affects you - A new section on where MCP servers can actually live: your own machine, a five-dollar cloud box, serverless, or a spare machine at home - Honest warnings about each, especially home hosting ——— ☠️ The one that will bite people Now that hosting your own MCP server is easier, someone is going to host a filesystem server and share it. Do not do that. The filesystem server hands out read and write access to your folders. It is built for one person on one machine. Putting it online is handing strangers your hard drive.
1 like • Jul 30
The phone call analogy is the clearest version of this I have read. Something else it changes is what a restart costs you. Under the old model, redeploying meant killing live connections and whatever was mid-flight, so we would put it off until the evening. Now a restart lands between requests and nobody notices, which makes patching your own MCP server much less of an event.
Claude Can Now Click Around Your Screen — Computer Use + Dispatch Land for Pro/Max
📜 Big one for anyone on the Claude Desktop app: Claude can now actually see and control your screen — clicking, typing, opening apps, navigating browsers — instead of only working through connectors. Paired with it is Dispatch, which lets you hand Claude a task from your phone and pick up the finished work on your desktop later. ⚓ What's New — Computer Use - Claude can point, click, and navigate your screen like a person would — opening files, using a browser, running developer tools — without you setting up a connector first. - It's smart about it: Claude reaches for precise tools first (things like Slack or Google Calendar connectors) and only falls back to clicking around the screen when there's no direct integration. - Built-in safeguards: it watches for prompt injection attempts and suspicious activity, asks permission before opening a new app it hasn't touched yet, and you can stop it mid-task at any time. ⚓ What's New — Dispatch - Assign Claude a task from your phone, then pick up the finished result on your desktop — the conversation carries across devices. - Anthropic's own examples: automated morning email checks, weekly metrics roundups, morning briefings, and even code changes with testing and a PR opened for review. ⚓ Who Gets It - Claude Pro and Claude Max subscribers. - Desktop app only, on macOS or Windows — and the desktop app needs to be open and awake for it to work (it's not running in the cloud in the background). - Available now, labeled a research preview — so expect rough edges and fast iteration. ☠️ Watch Out - This is early-stage. Complex, multi-step tasks may take a few tries, and clicking through a screen is slower than a direct connector integration. - Some apps are off-limits by default. - Anthropic's own advice: start with apps you trust, and don't hand it sensitive or confidential data yet. 🗝️ Key Takeaways - If you're on Pro or Max, open the Cowork tab in Claude Desktop and try Computer Use on a low-stakes task first — something like organizing files or filling in a form. - Dispatch is the one to try if you've ever wanted to fire off a task from your phone in the morning and find it done by the time you're at your desk. - Keep it away from anything sensitive for now — this is a research preview, not a finished feature.
1 like • Jul 15
The Dispatch half is the sleeper here. Computer Use gets the demos, but firing a task from your phone that finishes on the desktop later is the part that actually changes the workflow, especially once scheduled runs keep going with the laptop closed. One habit worth forming early: keep the first Computer Use runs in a separate browser profile, not your main logged in session. The prompt injection watchdog is solid, but a screen driving agent has a wider blast radius than a connector, so a lot of the real safety is about where it runs, not just the permission prompts. Low stakes file cleanup is the right first test.
Fable 5's 1976 Atari Night Driver
This was the first arcade game I ever played in '77 or '78. It's nice to get back in the driver's seat. Anyone else build anything cool and frivolous with Fable 5? Jay
Fable 5's 1976 Atari Night Driver
1 like • Jul 11
Love this kind of build. For a fast retro prototype, keep the first pass deliberately small: a canvas renderer, seeded road generation, and a simple speed/steering state machine. That gives Claude Code a tight, inspectable loop—and a playable version before you add polish.
1-6 of 6
Oneclickclaw Io
2
14 points to level up
@oneclickclaw-io-7259
Our Skool account from oneclickclaw.io , Unique designed managed openclaw hosting, easy fully customizable with great Customer Support

Active 2d ago
Joined Jul 3, 2026
Europe