📝 The Weekly Scroll: We’re in our building blocks-era
And you can be in yours, too
If you’ve ever looked at words like “API,” “pipeline,” or “integration” and quietly assumed they weren’t for you, this issue is for you. Half the people I’m about to walk you through don’t write code.
For the last six weeks, I’ve been working on articles for the Automations cohort of our Guest Post Program — a way to get people who are doing interesting things with Buffer onto our blog, in their own words.
Every article in this cohort is from someone who built something specifically with the new Buffer API. And since we’ve just launched the API, I thought I’d share some of the patterns I’ve been noticing across the people building with it.
Before we get into it, here’s what the cohort has built (so far):
Anthony turned an outdated Zapier setup into a custom job-board pipeline
Angie surfaces content ideas from her community’s Slack questions
Brandon’s product database now generates 21 posts a week across three platforms.
Joe’s Mac app turns a stack of Pokémon cards into eBay listings and Threads posts in one drag-and-drop.
Ben built a workflow layer between his idea capture and Buffer’s queue.
Violeta finds AEO gaps on LinkedIn and turns them into a week of posts.
Shivani turned her old LinkedIn posts into a searchable, repurposable library.
Fahad built an engine that stages trending topics inside Buffer. His piece is already live on the blog.
Miguel’s SwiftUI app turns his reading highlights into LinkedIn posts.
Martín wrote a script that auto-schedules daily posts for his word game, and the daily cadence accidentally sparked a community around it.
Some of them are engineers. Some have never written a line of code. Both groups built the same kind of thing. Now let’s get into how.
1. They built bridges, not new platforms
Every one of them connected two systems they were already using. Anthony’s CMS to social. Angie’s Slack to her content calendar. Joe’s eBay to his social following. Brandon’s product database to Instagram. Nobody was trying to build a new platform from scratch. They spotted a gap between two tools they already used, and built the small piece that connected them.
The thing that makes this newly possible for non-engineers is the tools that build the bridges. Angie’s runs on Gumloop. Shivani’s runs on Lovable. Neither of them wrote a custom backend — they described what they wanted and let an AI assemble the plumbing.
2. The trigger was always the same question
Every guest poster in the cohort started with a version of the same question: “I want this to behave slightly differently. How can I do that?”
When you’re asking that question every other day, that’s usually the sign you need something custom. The off-the-shelf tool worked but had a gap that kept tripping them up. So instead of waiting for a setting that might not be coming, they built the missing piece.
🛠 If you’re new to this, here’s the no-code stack the non-engineers used
If the word “API” still feels out of reach, this is the toolkit half the cohort actually used to build with the Buffer API without writing code:
Gumloop — drag-and-drop AI workflows on a canvas. (
Lovable — describe an app in plain English, get a working app back.
n8n / Zapier — the classic connect-anything tools, now with MCP support, so they speak to AI agents too.
Claude (with MCP) — talk to the Buffer API from your chat window. No code, just describe what you want.
Cursor — an AI pair-programmer that handles the actual code part for you while you stay in plain English.
The skill these tools ask of you isn’t code. It’s describing what you want clearly enough for an AI to execute.
3. AI did the assembly, but it was all driven by humans
None of these workflows is “full automation.” Every one has a deliberate human checkpoint. Anthony approves each job manually before the pipeline queues it. Angie reviews drafts inside Buffer before anything goes out. Joe sets the price himself on every card. Brandon spot-checks the week’s posts in iA Writer.
What that tells me is the question isn’t “how much can I automate?” It’s “which part of this work do I want to keep doing?” Everyone in the cohort answered that question for themselves before they shipped.
4. Voice was the part they obsessed over
Angie has an entire section in her piece titled “How I trained the AI to sound like me.” Brandon keeps a voice-and-tone Markdown file and spends more time on it than the code powering his automation.
Most of them spent more time tuning the voice than they did building the workflow itself. Every cohort writer eventually arrived at the same conclusion: the mechanics were (relatively) easy to build, but the voice couldn’t be perfectly replicated.
5. The most valuable output was time back, plus consistency
If you’re running something on top of a day job, a kid, a side project, or a volunteer role, this might be the outcome that matters most.
Nobody in this cohort got 10x more output from automating their stack. Brandon’s engagement didn’t skyrocket. But he went from impulse posting to daily posts across three platforms without losing his dev time. Angie got valuable time back. Anthony got five hours a week back and put them into the community he runs. Joe walks away from a stack of listings on a Sunday evening.
What these builds give you is time back, and a way to keep showing up without burning yourself down to do it.
What caught my eye most was how small the steps were to getting something functional. Each builder had something specific they kept tripping over, and they took a few hours to fix it.
“I built it with the Buffer API”, might sound intimidating, but the context of how little time it took and how little technical proficiency you actually need to get something working makes it seem far less daunting.
So, something to think about: where can you plug a gap in your current workflow?
API ALL THE WAY DOWN
You’ve heard from the builders. Now it’s your turn.
Buffer’s API is now live for everyone. And you don’t have to write any code to start.
The fastest way in for non-developers is to connect it to Claude or ChatGPT via MCP, then describe what you want in plain English and let the model handle the API calls.
Wire it up and get to testing and building. We can’t wait to see what you come up with.
WHAT’S IN MY SCROLL?
It’s been an API-heavy week, so most of this is API-adjacent.
First up a double feature, from the Buffer team:
How to use Buffer with Claude — the entry point for anyone who wants to wire up Buffer without writing code. Start here if “API” still sounds intimidating.
How to build with Buffer’s API — for when you’re past the chat-window stage and want to see what the API actually does under the hood.
Together, the two videos walk you through the full learning curve in about 20 minutes.
We also have case studies from different people who’ve built some very impressive things with the Buffer API to inspire your next build:
Ben Campbell on PostIQ
Shivani Shah on her LinkedIn Command Center
Writestack’s founder Orel Zilberman on cross-platform scheduling for Substack (!) creators
Marin Nedelev on managing 77 social channels in ten languages
That’s already a lot so I’ll sign off quickly with a request: if you’ve built anything interesting recently, share it in the comments! We’d love to champion more builders.
Until next week,
Tami
Sr. Content Creator