I'm building a multi-contributor project with three workspaces — writing, visual/creative, and business/marketing. Each contributor will have their own ICM workspace. I want shared global context files — character profiles, world-building, backstory, business overview — that sit above all three workspaces and feed into each one consistently. Two questions: 1. Can a workspace's CONTEXT.md reference markdown files that live outside its own folder — upstream in a parent directory? 2. Has anyone structured a shared GitHub repo where global files live at the root and individual workspaces pull from them? How did you handle keeping global files canonical when multiple contributors are updating their workspaces? I'm not sure how to go about this.. I'm definitely green around the edges 🌱 Thank you kindly!

I'm drafting a very old sci-fi novel of mine with Claude Code. Four scenes in. More excited about a creative project than I've been in years — and the reason isn't the speed. It's that the workspace is built to refuse. Setup: a folder called `writing-room`. Eight stages, from premise to compilation, each one a markdown directory the AI loads only when it's relevant. Compass, world, characters, structure, voice, writing, revision, compilation. The first rule, hardcoded in `CLAUDE.md`: > Before generating prose, always load `voz.md` and `padroes-prosa.md`. Without these two, refuse the writing task and ask the author to do Stage 05 first. Translation: the AI cannot draft a scene until I've locked in the voice. And `voz.md` was reverse-engineered from scenes I wrote by hand. The voice is mine. The AI only gets to extend it. There's also a file called `padroes-prosa.md` — 9 anti-AI-slope techniques. Verbalized sampling. Fragmentation. Character voice. Rare vocabulary. Every generated scene must apply at least 3, and the reviser uses the same file as a checklist. What this changes in practice: - I don't fight AI prose. I gate it. - Each stage loads minimum context. The AI doesn't drown in 200k tokens of worldbuilding to draft one scene. - After every scene, a `cronista` skill updates a canon file. Continuity stays cheap. - I'm the bottleneck on voice. I'm fine with that. The transferable bit, if you build with AI: The most useful thing your workflow can do is sometimes say no. Refusing to act without the right inputs forces you to produce those inputs — and that's where your taste enters the system. Without that gate, the AI averages you out. Toward the median sentence. The median plot beat. The median version of you. A friend of mine said that "in order to have a second brain, you need to have a primary working brain". I laughed: true enough. I wanted to build the gate first. Then let it write. And I'm loving it.

Tell us three things: 1. What you do (job, industry, student, career-changer, whatever) 2. What brought you to Clief Notes 3. One thing you're trying to figure out right now related to computing or AI I'll respond to every single one. And read each other's intros too because the person who's stuck on the same problem as you might already be in this thread. I'll go first I am Jake, I have been working in tech for 15 Years, building with Generative AI for 3 Years straight now! Excited to teach and learn! That's it. Simple, scannable, gives you data on who's joining and what they need, and keeps the feed clear for content that retains people past week one.

NLP Logix was founded in 2011 so if you wanna talk about being in AI before it was cool this company did it. Matt, the Chief AI officer sat down with me and chatted over what matters in the ai age. Check it out! (and go leave a comment on the YouTube video if you have time please!) They are looking at showing up to one of the next High Teas so keep an eye out for that announcement!

I can not stop running my mouth about the benefits of running 1 agent through these dynamic context / folder / markdown / skill trees. I'm ecstatic!!! What I learned building my first production AI workflow with ICM [newsletter builder] Where things broke — and what I learned Content accuracy doesn't self-police. The AI surfaced outdated content and presented it as current. We added a structured recency guardrail classifying content into three tiers — time-sensitive, forward-looking, and evergreen — each with its own freshness rules. Factual claims need a dedicated review pass. Vague language around eligibility, requirements, and designations slipped through editing. We added a formal fact-check pass with explicit criteria: verify claims against primary sources and flag anything requiring tight specificity. The build stage needs the right foundation. The first output was functional but not platform-native. Rebuilding it using our platform's actual template structure unlocked proper responsive behavior — and it deployed cleanly on the first attempt. [Sales Force template generation on the fly!!] Generated assets need a management workflow. AI-generated images were enormous files causing slow load times. Building a download, optimize, and host workflow into the stage documentation means every future issue handles it consistently. [Connected Higgsfield MCP] Updated our Sales Force template to the new HML syntax. -- Updating every reference in both the output and the skill documentation means future issues are built correctly from the start. The takeaway Every problem we hit became a documented fix that future runs inherit automatically. By the end of one session, the pipeline was producing accurate, platform-ready output with guardrails that didn't exist at the start. If you're building any recurring AI-assisted workflow, the investment in getting the stages right pays off.