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

Memberships

Clief Notes

31.6k members • Free

Mobility & Injury Prevention

211.8k members • Free

14 contributions to Clief Notes
Claude Projects doesn't accept folders
Claude Projects doesn't accept folders, only files. So the usual ICM layout (CLAUDE.md at root, /protocols, /reference, /templates, /examples) won't upload as is. What I did: - Flattened everything into a single folder. Renamed each file with its folder as a prefix, e.g. protocols/guardrails.md became protocols_guardrails.md. Keeps the origin visible in the filename. - Updated every internal reference inside the files so the File Map and cross-links point to the new flat names. - Pasted the body of CLAUDE.md into the Project's Custom Instructions field. CLAUDE.md has no special status in Projects (unlike Claude Code), so the orchestration rules need to live in the instructions box to actually govern behaviour. - Uploaded the other files as Project knowledge. Done. The File Map still works as a routing index. Early days but the structure holds.
0 likes • 9m
What is the benefit of claude projects over claude co-work?
🏆 WEEKLY COMP #4 RESULTS: THE AGENCY (CORRECTED) 🏆
I need to walk back this morning's announcement before anything else. @Ariel Ortiz is not a Premium or VIP member. The competition was Premium and VIP only, stated in bold in the original brief. That eligibility check should have happened before judging and it didn't. That's on me. The $325 and the Week 4 title go to @Ruby Sparks. The rest of the writeup stands. Repeating it here so the corrected results live in one place. ➖➖➖ 🏆 WINNER: @Ruby Sparks 🏆 🔗 Repo: https://github.com/sparkles-inc/agency-os 🌐 Live site: https://agency-os-tan-five.vercel.app/ Ruby was within an inch of taking it on the original call. The judging call between her build and Ariel's was the longest I've sat with on any comp. She earned this on the work alone. The Loom video she recorded for this submission is the cleanest voice work I've heard on any community submission. Full-on infomercial quality. Voice acting level. If you haven't watched it yet, go watch it. It's a master class in how to present a build. Beyond the voice: an animated handoff explainer on the companion site, a WRITEUP.md that compresses the entire submission into three paragraphs of clean argument, and a design philosophy made explicit. She refused typed schemas in favor of human-readable Handoff Cards with seven sections including a required Gaps field ("if Gaps is empty, you're not looking hard enough"). One continuous narrative thread runs through all 27 files of her repo. Ruby takes the $325. @Sonija Quinn will reach out to Ruby directly for payment details. Remember: in a world where AI is making it easier to build, it's the unique opinions and styles that become valuable. ➖➖➖ I thought last week was hard to judge. This was the hardest comp to judge so far. Not even close.
1 like • 58m
@Jake Van Clief you say the rest of the original writeup stands but you also deleted the original post so the top 5 (4 now) are gone 😢
0 likes • 29m
@Jake Van Clief oh. I apologise. User error on mobile the more button isn’t as obvious than the desktop 🤦‍♂️ ☺️ thanks
👾 What model do you use? May edition
Drop your vote everyone! Might be fun to do monthly, like this post if you agree or want more polls like this. Enter your answer below or in the comments 👇
Poll
16 members have voted
1 like • 20h
@Kevin Carrasco I understand completely. For some reason I have always had great success with sonnet. Every time I try another model it just doesn't behave as well as I would like. Cursors composer model is a close second, but i've never liked codex for code.
1 like • 20h
@Kevin Carrasco 100% My comment was mainly for code. Using the folder system I have tried out a few different models mostly on their free tier and they all perform pretty much the same for non-coding stuff. I.e. the weekly compo 4 submnission I tried out using gemini/codex/claude/mistral and they all did an amazing job
🏆 WEEKLY COMP #4: THE AGENCY 🏆
💰 $325 CASH 💰 That's a full year of Premium. Win this and your membership pays for itself. But the real prize this week isn't the cash. Keep reading. 📋 THE CHALLENGE You just got hired again. Meet Diana, owner of a 4-person boutique real estate team in Austin. 60-80 transactions a year, mostly residential, mix of buyers and sellers. 📎 Download the full client brief attached to this post. Short version: She doesn't want software. She wants a system she can teach her team to use in a week. Your job is to build the AI operating system for her team. This isn't one specialist. This is a small team of AI specialists organized into a multi-folder ICM architecture, with a clear handoff protocol between them. 🗂️ WHAT YOU'RE BUILDING Last week was one specialist. This week is a team of them. Required folders: 📍 00_orchestrator/ — The router. Where every request starts. Decides which specialist gets the job. 📍 01_lead_qualifier/ — First contact with new prospects. Captures intent, budget, timeline. 📍 02_property_research/ — Deep research on specific properties or neighborhoods. 📍 03_client_communication/ — Drafts emails, texts, follow-ups in the voice of the agent. 📍 04_transaction_coordinator/ — Handles the deal once it's live. Checklists, deadlines, document tracking. Each folder must include: - 📄 identity.md - 📐 rules.md - 💬 examples.md - 🔗 handoff.md (NEW for Week 4 — how does this folder pass work to another folder?) - Plus a root-level README.md explaining the architecture, the typical flow, and how to onboard a new team member. 🔥 WHY THIS ONE IS DIFFERENT Weeks 1, 2, and 3 were warmups. This is the comp where the work you ship genuinely starts to look like the real thing. The handoff protocol is the test. Anyone can build five folders. The hard part is defining what each one needs from the previous one and what it passes to the next one. That's where multi-agent systems actually live or die.
2 likes • 4d
@Andrew Baer Yes. Jump in and start doing. Even if its rubbish. I regret not joining the weekly challenges earlier. I was even thinking I didn't have enough knowledge to attempt this one. Get started, peruse the other entries, the learning is in the doing!
0 likes • 3d
@Ernie Cruz Great idea! There is no end to great ideas on this platform. the quick start readme would literaly be download and use, the orchestrator could just run a quickstart workflow to get the agent enrolled in the system. This is all insanity, so much more to learn! But to be fair its just headspace and realising what can and cant be done! Thanks for the comment!
Highest signal to nosie MD file length
What has everyone found to be their sweet spot when it comes to the line length of thier MD files? Wouldn't it be more efficient to have fewer lines in a markdown file but have them be higher quality? You would likely need to create more MD files, which will take more time. However from my understanding it's easier for LLMs to process higher quality MD files with more files than the other way around. I was wondering what everybody else thought of this.
2 likes • 5d
What @Niel de Kock said. Smaller files means smaller context. Small good quality content beats large files with poor content any day. Write for people not for agents. Why would someone read 10 paragraphs of fluff when 2 paragraphs of dot points will do the same thing! If something is not relevant for all requests, then put it in its own folder/file, and tell the LLM to go and read that only when it becomes relevant. The LLM for other requests will not be confused about context it doesn't need to be aware of, and you get more of your context window to do better quality work with less hallucinations. When the LLM recognises it needs that new context, it reads the file and now it magically knows more stuff and can change its behaviour accordingly.
1-10 of 14
Matt Paine
3
33points to level up
@matt-paine-4542
A

Online now
Joined Mar 31, 2026
Powered by