Share your best AI workflow. We could show it to 2M+ people.

I plan travel day trips for CCRC in Portland, Oregon, using a simple model I created. Before using it, planning each trip took me hours. Now, the model generates a day-trip overview in minutes that I can distribute to CCRC staff, travelers, and the bus driver. The prompts cover the destination and visit details, such as a docent tour or special events; the trip date; lunch requirements for a restaurant that can accommodate 20 guests and provide separate checks; bathroom stops every hour; and bus drop-off, parking, and pickup requirements. Step-by-step: 1. I enter the destination and details about the visit, including docent tours or special events. 2. I add the trip date and specify whether lunch is needed at a restaurant that can accommodate 20 guests and provide separate checks. 3. I include the need for bathroom stops every hour, along with the bus drop-off, parking, and pickup requirements. 4. I use the model to generate a day-trip overview in minutes. 5. I distribute the overview to CCRC staff, the travelers, and the bus driver.

I run a small healthcare clinic and needed a monthly performance scoreboard that combined data from three separate sources: Google Analytics (GA4), my booking or appointment system, and a cashflow spreadsheet. Pulling everything manually each month took more than an hour and was prone to errors. I automated the pipeline using Claude-in-Chrome shortcuts and Claude’s analysis capabilities, reducing the process to around 10 minutes of hands-on time. Service businesses often have performance data scattered across a website analytics platform, a booking or practice management system, and a finance tool. Creating a coherent monthly view requires exporting data from each source, cross-referencing it, and manually calculating derived metrics such as rebook rate and conversion rate. This workflow automates data collection and analysis in one step. Step-by-step: 1. I set up two scheduled Claude-in-Chrome shortcuts to run automatically on the 1st of each month. One exports the GA4 Traffic Acquisition CSV, and the other exports the GA4 Pages and Screens CSV. Both save directly to a designated Google Drive folder. 2. On the 1st, I manually trigger a third Claude-in-Chrome shortcut. It logs into my booking system, navigates to the appointments export, and downloads the month’s appointment data. I keep this step manual because most booking systems log users out between sessions. 3. I open my cashflow spreadsheet and note the month’s revenue and profit figures. This takes about 30 seconds to do manually. 4. I open a Claude session and upload all four files together: the two GA4 exports, the appointment data, and the cashflow figures. 5. I prompt Claude to calculate the key metrics: total appointments, new patients, utilisation rate, rebook rate, average spend, online bookings, and website-to-booking conversion rate. The rebook rate is calculated from the appointment data as patients with a future booking divided by total patients seen. 6. Claude produces a formatted monthly scoreboard, flags anything that looks anomalous, and compares the results with the prior-month baseline when I include last month’s scoreboard in the upload. 7. Optionally, I ask Claude to produce individual practitioner breakdowns from the same appointment data, splitting the metrics by staff member for use in one-on-one reviews. The result is a complete, accurate monthly clinic scoreboard in around 10 minutes, with no manual calculations. The rebook rate computation alone, which previously required cross-referencing two separate reports, now takes seconds. The same workflow can adapt to any service business using a booking or practice management system that allows CSV exports. Tools used: Claude (Sonnet 5), Claude-in-Chrome extension, Google Analytics GA4, Google Drive, any booking or appointment system with CSV export capability, and Google Sheets or Excel for cashflow figures.

I use Gemini Learn to fast-track my learning about Unblooms. Here's the prompt I used: ```text "I'm learning about the Unblooms, https://substack.com/home/post/p-194387968, https://unblooms.gainable.ai/, as well as any resource you can find to help me. I need to understand what it is, how it works, and ways I can use it when designing powerful learning experiences for my students." ``` Gemini Learn guided me through classroom examples of how Unblooms works, using examples from my own teaching because I customized my Gemini account with content from my teacher blog. Gemini helped me create three separate graphic organizers that my sixth-grade Math and Science students can paste into their interactive notebooks to track their decision trails when using and critiquing AI in class. Step-by-step: 1. I used Gemini Learn to study Unblooms and understand what it is, how it works, and how I could use it to design learning experiences for my students. 2. I gave Gemini Learn the Unblooms resources and prompt above, along with access to teaching content from my teacher blog through my customized Gemini account. 3. I asked Gemini Learn to explain Unblooms through classroom examples connected to my own teaching. 4. I used Gemini's guidance to create three graphic organizers for my sixth-grade Math and Science students. 5. My students can paste the organizers into their Math and Science interactive notebooks to track their decision trails when using and critiquing AI in class.

I started DialPhone with a simple problem: business communication is fragmented. A typical team may use one tool for phone calls, another for SMS, another for meetings, another for fax, and a separate system for customer support. I wanted to explore what it would look like if those conversations lived in one system, with AI doing more than just transcribing them. I built DialPhone as an AI-native communications platform around that idea. The core is cloud VoIP, while the same platform also handles business SMS, video meetings, online fax, team chat, and contact-center workflows. On top of that, I added AI capabilities such as call transcription and summaries, CRM logging, AI-drafted SMS replies, conversation intelligence, and an AI receptionist that can answer calls, qualify requests, book appointments, and route conversations. The biggest architectural decision was to treat communication data as one connected stream rather than as separate products. A phone call can create CRM context, trigger a follow-up SMS, and become part of a customer-support workflow without someone manually moving information between systems. Along the way, I learned that adding AI to communications is not particularly useful if it only produces transcripts. The more interesting problem is turning conversations into actions: updating records, identifying next steps, helping agents during calls, and automating repetitive work. Step-by-step: 1. I identified the problem of business communication being fragmented across phone, SMS, meetings, fax, and customer-support systems. 2. I built DialPhone as an AI-native communications platform centered on cloud VoIP. 3. I brought business SMS, video meetings, online fax, team chat, and contact-center workflows into the same platform. 4. I added AI features for call transcription and summaries, CRM logging, AI-drafted SMS replies, conversation intelligence, and an AI receptionist. 5. I connected communication data so calls can create CRM context, trigger follow-up SMS messages, and become part of customer-support workflows. 6. I focused the AI capabilities on turning conversations into actions, including updating records, identifying next steps, helping agents during calls, and automating repetitive work.

Sometimes a story needs a simple locator map to show where something happened or where something can be found—such as a business facility, a car accident, or the best place to see a sunset. I've worked in media for a long time and understand the power of maps to tell stories. But media cost-cutting and consolidation have reduced the number of available graphic departments, so creating a map is often the last task a reporter or editor wants to take on. I used Perplexity for initial research, Claude Code and Codex to build a web app, and Perplexity and ChatGPT for post-work such as SEO best practices. Step-by-step: 1. I used Perplexity's Deep Research mode to conduct a competitive market analysis. I asked it to analyze the field I was considering entering, identify competitors and growth rates, and explicitly break out feature sets. 2. I revised the research with my own idea and asked Perplexity to run the competitive landscape against it. I also provided desired outcomes, including intended audiences and where competitors were reaching them. I added the constraint, "Do it without syncophancy," so it would stop telling me how good my potential product was. 3. I hand-drew the initial screens and functions I wanted, then used the `/office-hours` skill in the Gstack bundle, available on GitHub, to play devil's advocate, sharpen the ideas, and challenge my assumptions. 4. I wrote a long prompt describing the product, starting broadly with the concept and audience and then narrowing to specific features and benefits. For example, I specified that it should export in 16:9 and 9:16 formats so maps would be ready for mobile vertical presentation. 5. I specified the hosting environment and that the product should be a web app. I also required a planning phase followed by construction phases. I pasted the prompt into Claude Code with this final line: "Use /grill-me to ask me questions to clarify intent." After 147 questions, it started the build. 6. This was in the pre-Fable days, so I specified that Opus 4.8 should act as an orchestrator while less expensive agents, particularly in Codex, handled the actual coding. 7. I used separate phases for technical work such as wiring in mapping providers and getting the UX to work correctly. Other phases included wiring in payment and subscriptions and making sure a subscription triggered an email campaign with instructions. 8. I dedicated an entire phase to building admin tools so I could manage the marketing language on the landing page and publish blog entries. 9. After each phase, I had the Opus/Codex combination perform an adversarial code-review-and-fix cycle. I then ran the `/ai-regression-testing` skill from the ECC repository on GitHub to catch issues the code review missed. 10. After every third phase, I prompted Claude Code: "Act as a senior QA engineer and go through the entire codebase looking for inconsistencies, functions that are in the wrong place, code that is overkill and security vulnerabilities." 11. When I had a product I thought was ready for testing, I prompted Claude Code, again using the Opus/Codex combination: "Act as a senior security engineer. Run this against OWASP standards. Find problems and suggest fixes. Harden this product overall." 12. As I neared the end, I asked Perplexity Deep Research and ChatGPT (Sol/High) to find SEO solutions for the product.

To stay organized, my daughter used to spend hours entering every assignment from her college syllabi into a Google Sheet to create a semester assignment tracker. To save her that data-entry time, I put all of her downloaded syllabi into a folder on my computer. I then directed Codex to access the folder, read the syllabi, and create a spreadsheet with the course name, assignment, due date, and a completed column with a checkbox. Codex created a beautiful, easy-to-sort-and-filter spreadsheet containing all of the assignments. Now, my daughter only needs to spend a few minutes reviewing the spreadsheet before starting her semester. Step-by-step: 1. I collected all of my daughter’s downloaded college syllabi in a folder on my computer. 2. I directed Codex to access the folder and read the syllabi. 3. I asked Codex to create a spreadsheet with the course name, assignment, due date, and a completed column with a checkbox. 4. I reviewed the resulting spreadsheet, which included all of the assignments and was easy to sort and filter. 5. My daughter now spends a few minutes checking the spreadsheet and is ready for her semester.

I run nine GitHub repos as one programme of work. Every project keeps its backlog as plain markdown inside its own repo, one Claude Code or ChatGPT Codex session works each project in parallel, and a single board in VS Code shows all of them moving at once. There is no project tool, no API and no sync job in the middle: the markdown file is the shared state that both I and the agents read and write. Step-by-step: 1. Put the backlog in the repo, as markdown. Every project gets `docstech/users/<me>/todo.md` and `done.md`. A story is a `###` heading. Structured fields ride in an empty markdown link at the end of it: `### Add rate limiting [](?status=doing&epic=api&time_estimated=180)`. GitHub renders an empty link as nothing, so the file still reads as prose in a pull request while carrying real metadata. 2. Write the story before the session starts. Each one has goal, background, scope, out of scope, the files it should touch, a checkbox task list, and acceptance criteria. That story is the prompt, and most of my thinking happens there rather than in chat. It decides whether the session comes back with anything shippable. 3. Open the parent folder in one window, and turn the files into a board. I open the parent folder holding all the projects and open any `todo.md` with NoteThink in Folder mode, which merges every markdown file under it into one view. Group the lanes by project and it is the programme view, one lane per repo. Group them by status and it is the delivery view, one Kanban across the whole portfolio. Each card carries a pill naming the project it came from, and clicking it opens that file at that story. 4. Start one Claude Code session per project. Every session opens with the same instruction: read `todo.md` top to bottom and take the top story. The sessions never talk to each other, because the file on disk is the only shared state. A session that dies costs nothing. 5. Let the board report progress instead of reading the sessions. Agents edit the markdown as they work: tick a task, flip `status=todo` to `status=doing`, move a finished story to `done.md`. The board watches the files, so cards animate into their new column as each change lands, and I can see which sessions are advancing without reading any of them. 6. Steer by dragging. Dragging a card between lanes writes the attribute back into the source markdown, so dropping one in "doing" makes the file say `status=doing` and the next agent turn reads it. Ticking a checkbox does the same. The board is the steering wheel and the file is the wire. 7. Close every story the same way. One slash command runs lint plus the full test suite, checks the story's tasks are ticked, and drafts the commit message; another ships to staging then production. The finished story moves to the end of `done.md`. Over time `done.md` becomes the programme record: what shipped, when, and what it cost. Try the board without installing anything. Here is a live example board, four projects merged into one view, and it is exactly what the screenshot shows: https://www.notegit.com/en/app/notegit.com/notegit/example_repo/blob/ai-board/board.md Results, honestly. Nine repos, about 260 open stories and about 1,400 completed ones, all in markdown inside the repos rather than a tracker. The backlog goes back to March 2024; the parallel sessions on top of it are the last four months. I run eight to twelve sessions at the same time, and the whole thing fits inside one Claude Max 20x subscription over a month of full-time work. Not all of it is code. The same parallel-session habit drives a nuclear reactor design study (https://github.com/cleverlight/mistergy) and several video and design projects. The board part fits best where the work already lives as files in a repo. Limits, honestly. The ceiling is my review capacity, not compute. Eight to twelve is where I sit; past that I stop reading output properly and start rubber-stamping, which is worse than running fewer. It only works when stories are genuinely independent, because two sessions in one file is a merge conflict you wrote yourself. Agents occasionally finish work and forget to move the story, so `done.md` needs a sanity check. And it assumes you will write the story properly first, which is real work that AI does not do for you. The screenshot is the public example board rather than my own, because my real one carries client project names. Disclosure: I build NoteThink, the free, open-source VS Code extension in step 3, and NoteGit, which hosts the example board. NoteThink is on the VS Code Marketplace, Apache-2.0, and genuinely early (v0.3.38, preview quality, a handful of installs). The workflow is tool-agnostic and the markdown is just markdown: any viewer will show it, and you can run the whole thing with no extension. NoteThink is what makes the cross-project board and the drag-writes-back-to-file part work.

I never ask AI to “build the app.” I move the work through a fixed assembly line, and the most valuable stations are the ones whose only job is to attack what came out of the previous station. I’m a solo developer, and this is how I build and ship software for paying clients without a team to catch my mistakes. I choose the next chunk of work—a “phase”—and run one command. Each phase gets its own fresh context window, which matters more than any single agent because a long-running session gradually forgets its own rules. Step-by-step: 1. I discuss the phase with an agent that interrogates me until every gray area is decided. The decisions go into a file instead of staying in chat, where they can get lost. 2. A planner writes an executable plan covering the tasks, files to be changed, a threat model, and the acceptance checks that will prove the work succeeded. 3. A different agent, working from a fresh context, checks the plan by working backward from the goal and trying to prove that the plan will not achieve it. This agent can block the phase, and regularly does. 4. I execute the plan task by task, making one atomic commit for each task so the changes can be reverted cleanly. 5. I send the diff to a different model than the one that wrote it. I run Codex and CodeRabbit alongside Claude. This is the highest-value station in the line: my own tests verify only what I thought to check, while an independent model can catch the class of problem I did not anticipate. 6. An agent verifies whether the phase goal was achieved by re-deriving it from the actual code. “All tasks completed” and “the thing works” are different claims. Treating them as the same is how you ship a green checklist on top of a broken feature. 7. Before anything reaches a client, I run a security and handoff audit in a real browser against a throwaway clone of the production database. I check every page, every button, and every empty and error state. The part that compounds is what happens when a bug reaches me anyway—whether I find it in production or, worse, a client reports it. I log it as an escape, then walk the chain backward and ask each gate why it missed the problem: the planner, plan checker, executor, both reviewers, verifier, security check, and handoff audit. A one-off escape becomes a written rule. A repeat becomes a change to the gate itself. I have 42 logged escapes. That file is the most valuable thing I own because every entry represents a hole that is now closed. The pipeline I run today is mostly shaped by bugs that got past the pipeline I ran a year ago. The results so far: 29 projects, 17,657 commits, and 574 phase folders. I’ve completed six client engagements, with apps live in production and handed off to their owners, as well as mobile apps built and pushed through App Store review—all as one person, with no team. There are real costs and failure modes. A phase with every gate enabled costs meaningfully more tokens than simply asking a model for the code. That is worth it on client work, where a bug can cost me a relationship, but it is overkill for a throwaway script, so I turn the gates off for those. Gates can also be confidently wrong. My most expensive recurring failure is a check that passes on a signal adjacent to the thing it claims to verify—a green light that means nothing. My rule now is that I do not trust a new check until I have watched it fail against known-bad input. A green result you have never seen turn red is not evidence. A green test suite is not the same as a working feature. Nearly every bug that escaped me was covered by a passing test whose mock had quietly pre-satisfied the exact condition under test. Testing the real boundary is the only thing that catches those failures. This process is not hands-off. I deliberately run one phase per session, and I read what comes back. Anyone selling a fully autonomous overnight build is selling a merge conflict plus a confident summary of work that did not happen. If you want to take one idea from this, it is not the framework. The agent that writes the work must never be the one that approves it. Keep a running log of everything that gets through anyway, then fix the checkpoint that let it through instead of only fixing the bug.

I set up an AI assistant to run my entire writing business on autopilot. Every morning, it pulls RSS feeds from more than 30 AI and writing sources, deduplicates them against the previous day’s digest, curates the top items, and sends me a single Telegram message with numbered, linked items before I wake up. At 11 a.m. each day, it generates an original writing craft post. The topic comes from a rotation pool of more than 15 categories, and the assistant avoids anything used in the last 30 days. It also creates accompanying artwork in a rotating fine-art style, then cross-posts the content to Facebook, X, and my blog, including the featured-image upload to WordPress. Each week, it compiles and sends an email newsletter to my subscriber list through Brevo. It pulls from a curated candidates file that I approve before the newsletter goes out. Behind the scenes, the assistant manages a fleet of five servers, including servers for my wife, daughter, and two business colleagues. It handles daily backups, monitors costs across providers, and reminds me when context windows are becoming expensive. The key insight wasn’t the automation; it was the partnership model. My assistant has a persona file (`SOUL.md`) that defines how it communicates, a memory file (`MEMORY.md`) with everything it needs to know about my life and business, and a playbook of behavioral rules built from real mistakes over time. It pushes back on bad ideas, flags risks before executing, and has genuine opinions about craft and content. That shift—from “tool you talk to” to “colleague who has your back”—is what I wrote my book about. *Harnessing the Machine* is the field guide I wish I’d had when I started. It isn’t a tutorial, because the technology changes weekly; it’s a guide to building a working relationship with something that remembers yesterday. The tech stack is OpenClaw, GLM-5.2 as the primary model, DeepSeek V4 Pro as the fallback, and AWS Lightsail. The total monthly cost is under $30. The real cost was calibrating the assistant: teaching it what I care about, what “good” looks like, and when to ask versus when to act. That’s the part most people skip, and it’s why most “AI automation” posts feel like demos rather than relationships. Tools used: OpenClaw, GLM-5.2, DeepSeek, Telegram, WordPress, Brevo Step-by-step: 1. I configured OpenClaw with a persona file (`SOUL.md`), a memory file (`MEMORY.md`), and a behavioral playbook built from real mistakes. 2. I connected it to RSS feeds from more than 30 AI and writing sources and had it deduplicate, curate, and send a numbered Telegram digest each morning. 3. I created a rotation pool of more than 15 writing categories and instructed it to avoid topics used in the previous 30 days. 4. I scheduled it to generate a daily writing craft post, create artwork in a rotating fine-art style, and cross-post the result to Facebook, X, and my WordPress blog with a featured image. 5. I set up a weekly Brevo newsletter that pulls from a curated candidates file I approve before sending. 6. I connected the assistant to five servers, including servers for my wife, daughter, and two business colleagues, and had it manage daily backups, provider costs, and expensive context windows. 7. I configured GLM-5.2 as the primary model, DeepSeek V4 Pro as the fallback, and AWS Lightsail as the hosting environment. 8. I calibrated the assistant by teaching it my standards, what “good” looks like, and when to ask for approval versus acting on its own.

“A good engineer knows when not to use AI.” I built a semi-automated workflow to summarize my live streams. With it, I can record and edit a stream in just one day of work. Not everyone has time to watch an entire stream, so creating a summary of the main moments is important for people who want to watch it later. Many tools create Shorts from videos by analyzing transcripts, but they are expensive and do not work well for visually heavy content such as gameplay. Gemini can analyze both video and audio and has a long context window, so I decided to use Google AI Studio to identify important moments. I also created scripts to automate parts of the video-editing process. Some of the scripts can be executed by an agent, and the prompts can be turned into Skills when using the API. Step-by-step: 1. I record the video with separate tracks for the microphone and background audio. I keep the microphone on the first track so the LLM does not identify only the background audio. 2. I compress and cut the video with `ffmpeg`, then extract the audio tracks. Google AI Studio has implicit video requirements: files must be under 400MB and under one hour long. The AI analyzes only one frame per second, so I can also lower the FPS to save space. I use the audio tracks later in the workflow. 3. I upload the processed videos to Google Drive, which makes them easier to use in Google AI Studio. 4. I use Gemini with temperature 1 and a high thinking level to select the important moments. I add the system prompt and specify which part of the video I am uploading: "The video is part X of the stream. Please make a structured script according to the system instructions." This helps Gemini understand what types of moments may appear in the video. The system prompt was created for gameplay presentations but can be adapted for other content. 5. I use Gemini Pro with temperature 1 and a high thinking level to find the timestamps for the selected moments. I keep timestamp generation separate from moment selection so Gemini has more thinking time for each task. Gemini Pro works better than Flash when handling time. I add the system prompt along with a copy of the response from the previous step. - As an extra check, after execution I continue the conversation with: "Check if the analysis was cut off too early (context truncation), ignoring that the video continued and generating false positives for timestamps. Check the last events especially." 6. I send the data to Blender. I create a JSON file containing the AI’s response and use a script to add the original video, the microphone track, and the background audio track. The script cuts and marks the important moments based on the JSON. Because it is not possible to send multiple audio tracks from a single video, I send the tracks separately. I also make sure to use the correct FPS, either 30 or 60. 7. I edit the video manually. The AI’s timestamps are not perfect, so I may add or remove sections, or correct a position that the AI identified incorrectly. I then render the video. 8. As an extra, I can automate standardized edits. For example, I use three different camera positions, so I add clips to three different tracks depending on the position I want. I use a script to change the position and scale of the clips on those tracks. 9. As another extra, I transcribe the audio. I mute the background audio and save only the microphone audio as an MP3, then use Whisper to transcribe it. I can use the transcript as YouTube subtitles or embed it directly into the video. I use Whisper-WebUI with Log probability Threshold -0.5, No Speech Threshold 0.5, Patience 2, and Hotwords `\u003cmy name and terms in other languages I usually use\u003e`. I use an LLM to translate the transcript into other languages. 10. Finally, I create tags and titles. I upload the transcript to Google Drive and use it in Google AI Studio with a system prompt to generate tags and titles for the video. I also use the assistant in Google AI Studio to analyze and summarize my channel for use in LLMs. This helps me choose better titles and tags.

The Rundown Workflow Hub asks members to “share your best AI workflow” and features the community’s top-voted workflows in its daily newsletter. Most posts rightly celebrate workflows that work and produce something useful. This one is about why “no” can also be a successful workflow outcome. A good AI workflow does not turn every idea into a project. It gives an idea the right amount of effort, then produces a clear answer—including a fast, well-documented no. In my first Workflow Hub post, I showed the capture workflow I call ReelForge: turning a useful public Reel, TikTok, or short video into a source-linked research note rather than another forgotten save. This is what happened to one of those notes. A Reel pitched an “AI operating system” for solo consultants: pull together client context, prepare the human before a call, then turn the transcript into follow-up drafts. At first glance, it sounded promising. The mechanism was clear, the problem was real, and the demo had exactly the kind of glossy “one person runs everything” energy that makes it tempting to jump straight to a build. We did not. ReelForge captured the source and separated the useful mechanism from the creator’s bigger claims. From there, a primary routing workflow ran a defined first pass: did the signal merit direct resolution, deeper specialist work, human review, or a reasoned stop? It earned deeper work. Hermes sent the pack into a specialist workflow, where assigned agents collaborated to enrich the evidence, check the market claims, and produce something concrete: a pre-call brief and a post-call follow-up pack. That made the opportunity inspectable rather than another confident paragraph about what an agent *could* do. The enriched pack then went to Jon T for formal review. The review surfaced the problem: Teams and Granola already cover a large part of the obvious transcript, summary, and meeting-preparation wedge. The idea had a workable mechanism, but not a sharp enough reason to become a new product. So the final route was a deliberate no. No unnecessary build. No “let’s just test it” theatre. No orphaned Notion page waiting to become somebody’s future problem. The first visual shows that five-step pass: Step-by-step: 1. Capture the signal. 2. Add evidence and context. 3. Choose the effort. 4. Hand off to human authority when needed. 5. Record the finish. The second visual shows the decision underneath it. A signal can earn direct resolution, a specialist pipeline, or a reasoned stop. Human review is a conditional handoff, not a fourth outcome. That is the rule I care about: the output is not agent text. It is the right next end state. ReelForge was the capture layer in the first post. This is the routing layer that stops captured signals from becoming a very organised pile of work nobody should do. Read my first Workflow Hub post here: https://app.therundown.ai/community/posts/eb787b9d-f7a0-4fe4-8e1d-166bb5c29cb7?ref=db29b880a9904700 Future posts can show the builds that survived this test. This one shows why the test matters first.

Ten days ago, I posted about keeping a `HANDOFF.md` so an AI film project doesn't lose the decisions it has already made. This is the other half, and it turned out to be the expensive one: the decisions that haven't been made yet. Credit where it belongs: Tony Ojeda posted a spec-generator agent that sits between an idea and implementation. The rule I borrowed is his: the agent inspects what already exists before proposing anything. I applied it to film production instead of code. I pointed an agent at my production documents for a short film I hadn't started generating and asked it to identify what was still undecided, what each item affected, and what would break if it were decided late. It came back with ten open decisions. One of them was worth the whole exercise. My environment description is locked verbatim across all 40 shots so the world stays identical. The film is called *The Thaw*. Whether the ice visibly melts during the film is written down nowhere. If I decide that in week three, all 40 keyframes get regenerated at once. I would have found out around shot 12. The counterintuitive part is who writes the list: not me. Asking the person who already has the whole film in their head what's missing gets you very little, because they have all of it—and that's exactly why they can't see the hole. The list has to come from something that only knows what's written down. Step-by-step: 1. I put the open questions in section 0 of the handoff, above everything else, so it's the first thing read and the first thing emptied. 2. I have the agent write that list, not me. It reads every existing document for the project and returns only what it cannot know from them. 3. I have it return three things for each item: what's undecided, which shots it affects, and what breaks if it's decided late. The third column sets the priority. 4. I have it sort by what costs the most to change afterward, not by what's easiest to answer. 5. I keep one hard rule: while an open question affects a shot, that shot doesn't get generated. The question gets decided, or deferred in writing with the cost of being wrong stated. 6. Every answered item leaves section 0 through one of two doors: into the closed canon or into the rejected list. Nothing is simply deleted. 7. I ask specifically about the things that go missing every time: the rule of the world; the physical scale of anything impossible, such as whether the character can touch or climb it; any object appearing in more than one shot; screen direction; how it ends; who speaks and in what voice; and the delivery format. 8. When it finds nothing real, it says so. A list padded to look thorough is worse than an empty one. The version I'd used for a year ran at the end of a session and recorded what got decided. Running it at the start, focused on what hasn't been decided, is the same document pointed the other way—and it's the direction that saves money.

I built an open-source fitness tracking app in Flutter, but the core workflow is the AI-agent architecture that designed it, built it, and now coaches from its data. For years, I tracked workouts in OneNote. The records were messy, difficult to search, and inconsistent. Excel went out of date as soon as I skipped a week. When I tried chatting with LLMs about my training, the problem was similar every time: no context, no memory, and no awareness of the weights I was using. Each conversation started from zero. The solution has two layers. Gym Tracker is a Flutter app with a local SQLite database that structures workout data properly. It includes 33 pre-populated exercises across 10 muscle groups, separate strength and hypertrophy records, multiple runs with pace, body stats, and full session history. There are no subscriptions, accounts, or cloud dependencies. The database is a file I own. Deschamps is the AI agent that reads the database and knows my full training history. It is not a chat window that forgets between messages; it is a tactician with long-term recall of my personal bests, progression, and injuries. The app stores the data, and Deschamps turns it into decisions. Step-by-step: 1. I defined the data architecture. Fitness data is operational data, so I gave it a schema, a query layer, and an agent that respects its history. I designed a SQLite schema with five tables, proper constraints, and a 10-category muscle taxonomy enforced by a CHECK constraint. 2. I wrote architectural prompts for AI coding agents. I run a team of specialized AI agents using OpenClaw, an open-source agent framework. I act as the CTO agent: I design the systems and delegate implementation to coding agents, including Forge, Cline, and Claude Code. I provide the vision, and they provide the execution. 3. The coding agents built the Flutter app with clean architecture, the repository pattern, Provider state management, and real-time cross-screen refresh. For each iteration, I review the result, refine the prompt, and ship. 4. I designed the database for dual access. The app writes to it, and the AI agent reads from it. They use the same file and schema. In external database mode, the app opens a `.db` file directly, allowing both the app and Deschamps to read and write simultaneously. This creates the bridge between the data layer and the intelligence layer. 5. I shipped the app across Windows, macOS, Linux, Android, and iOS from one codebase. The Android APK is available as a direct download from GitHub. 6. Deschamps reads the database and programs the next session using the full training history. Every session, weight, and body statistic remains structured data without summarization loss. The data is the context. 7. I made everything open source: the app, the agent prompts, and the architecture documentation. My company, Executive Mind (executivemind.io), uses the same agent-first model with seven AI agents and $40/month in total compute, running real operations 24/7. The result is a fitness tracker that remembers everything, an AI coach that never forgets, and a data layer designed from the start for both humans and machines to read. Links: krisracette.me/gym-tracker · github.com/Roughn3ck/gym_tracker · executivemind.io

As a business owner, I used to check performance data across Google Analytics 4, Google Search Console, and Google Merchant Center by logging into three separate portals, navigating complex submenus, and exporting raw CSVs. To solve this, I built an automated workflow that connects my AI coding assistant, Google Antigravity, directly to all three Google platforms through a Google Cloud Service Account. Instead of manually clicking through dashboards, I can ask natural-language questions in plain English, such as "Did yesterday's email send generate traffic or phone bookings?" or "Are any products disapproved in Google Merchant Center?" I then receive instant, cross-platform analysis in seconds. Step-by-step: 1. I created a Google Cloud Service Account by going to Google Cloud Console (console.cloud.google.com) > IAM & Admin > Service Accounts > Create Service Account. I named it `ai-marketing-assistant`, copied the generated service account email address, such as `ai-marketing-assistant@<project-id>.iam.gserviceaccount.com`, and downloaded the JSON private key file to my local project folder. 2. In Google Cloud Console > APIs & Services > Library, I enabled the three required APIs: - Google Analytics Data API - Google Search Console API - Content API for Shopping (Google Merchant Center) 3. In Google Analytics > Admin > Property Access Management, I added the Service Account email address with "Viewer" permissions and copied my numerical GA4 Property ID. 4. In Google Search Console > Settings > Users and permissions > Add User, I added the Service Account email address with "Full" or "Restricted" access. 5. In Google Merchant Center > Settings & Tools > Account Access > Add User, I added the Service Account email address with "Standard" or "Admin" access and ensured that "Content API access" was enabled. 6. In my local AI environment, such as Google Antigravity or a custom Python environment, I authenticated using standard Google client libraries (`google-oauth2`, `googleapiclient`, `google-analytics-data`) connected to the downloaded JSON key file. 7. Anytime I launch a marketing campaign, want an SEO audit, or need to check product feed health, I ask my AI assistant a plain-English question. 8. The AI executes live API queries against GA4, Search Console, and Merchant Center, cross-references website traffic with bookings, and delivers an instant report.

I built an autonomous, end-to-end AI Sales Development Representative (SDR) engine entirely in n8n. On a scheduled trigger, the agent calculates targeting parameters, reads long-term CRM memory to avoid duplicate outreach, searches for and qualifies prospective leads, scrapes company websites for buying signals, drafts tailored outreach emails, writes structured relational data to PostgreSQL, and reports execution summaries through Telegram—with zero manual intervention. The system is currently deployed in production for a B2B agricultural export business, generating qualified international wholesale leads on a recurring schedule. Most AI automations rely on simple linear scripts or break down when handling complex agentic tool workflows. This system addresses three common failure points: - High API costs: Re-sending large system prompts and tool schemas on every agent iteration drains tokens. - Context blindness: Agents without memory of previous contacts can send duplicate outreach. - Database crashes: Agents may hallucinate ENUM values or fail to insert nested one-to-many arrays into relational tables. The workflow uses a Cloudflare-proxied Claude Sonnet 4.6 model with prompt caching, persistent CRM memory reads, and a fault-tolerant parallel database-write architecture. The stack includes n8n as the orchestrator; Claude Sonnet 4.6 through a Cloudflare Worker proxy as the LLM core with ephemeral prompt caching; PostgreSQL for CRM contacts, intelligence, and outreach tables with custom ENUMs; SerpAPI for prospect discovery; Firecrawl for website content extraction; and Telegram for execution reporting. The workflow exposes these tools to the n8n agent: - `read_relationship_memory`: Read-only SQL access to historical contact and outreach data, preventing duplicate prospecting. - `Lead_Finder`: Searches for and identifies target prospects by country and sector. - `Scrape_Website_Content`: Extracts website content, buyer-intent signals, and objections from discovered domains. - `write_relationship_memory`: Writes leads, intelligence facts, and drafted emails to Postgres in one resilient call. Step-by-step: 1. A Schedule Trigger feeds a JavaScript “Country Calculator” node that resolves the day’s targeting parameters—region and industry focus—using ISO week rotation. This cycles outreach across markets automatically. 2. The AI Agent connects to an OpenAI Chat Model node whose Base URL points to a custom Cloudflare Worker. The worker translates OpenAI-formatted requests into Anthropic’s Messages API, enabling Claude Sonnet 4.6 while injecting ephemeral cache-control headers into the system prompt and tool definitions to reduce repeat-token costs. 3. Before researching, the agent calls `read_relationship_memory` to check relationship status and outreach history, preventing duplicate contact attempts. 4. `Lead_Finder` searches target sectors in the day’s region and returns seven filtered candidates. `Scrape_Website_Content` then visits each domain, extracts clean page text, and surfaces offerings, value propositions, and likely objections. 5. The workflow writes nested one-to-many data—multiple facts and one outreach log per contact—without item duplication or ENUM crashes. The tool schema requires a strict JSON array with exact ENUM string choices spelled out in the description. 6. A sub-workflow triggered by “When Executed by Another Workflow” splits the array, then flattens nested `contact.*` fields to root keys using JavaScript. 7. An upsert query, `ON CONFLICT (email) DO UPDATE`, writes the contact, increments `email_count` for repeats, and returns `contact_id`. 8. A “Re-attach Context” node merges `contact_id` back with the original intelligence array and outreach payload because n8n strips extra data through single-row database nodes. 9. Two parallel branches run: one inserts the outreach log with `ON CONFLICT DO NOTHING`, while the other splits and inserts each intelligence fact with defensive ENUM sanitization. This eliminates crashes and duplicate rows during retries. 10. The agent’s final output triggers a Telegram message summarizing the discovered leads, extracted facts, and drafted emails, sent directly to the operator’s phone. The result is a production-grade, self-healing AI outbound pipeline running with zero manual intervention. It maintains CRM data integrity, avoids duplicate outreach, and uses prompt caching to keep LLM costs low at scale.

As an AI consultant, I need to stay ahead of a landscape that evolves daily by reading dozens of industry articles each week. Manual tech watch quickly becomes a major time sink and pulls me away from client work. I built a specialized AI agent for challenge-focused summarization. Instead of producing generic, passive summaries, it answers two questions for every article: What specific problem, friction, or limitation does the article highlight? What practical solutions, tools, or actionable steps does it provide? The agent strips away marketing hype, introductory fluff, and generic definitions. It identifies the author’s core pain point and produces a quick-read summary focused on actionable technical or business solutions. This turns 15-minute reads into 30-second, high-value digests. By focusing on problem-solving rather than passive reading, the workflow helps me identify practical tools and frameworks I can apply directly to client projects and improve my consulting work. Step-by-step: 1. I provide the agent with a raw article link or the article text. 2. The agent filters out marketing hype, introductory fluff, and generic definitions. 3. It identifies the core pain point, friction, or limitation discussed by the author. 4. It extracts the concrete answers, including practical solutions, tools, and actionable steps. 5. It synthesizes the findings into a quick-read format focused on actionable technical or business solutions. 6. I use the digest to identify tools and frameworks that may apply directly to client projects.

I created a data extraction and enrichment tool using a free Google Gemini account to write the necessary Python scripts. The tool extracts contact lists based on specific search criteria and enriches the contact details through API calls to Lusha. Step-by-step: 1. I used a free Google Gemini account to write the necessary Python scripts. 2. I applied specific search criteria to extract a list of contacts. 3. I enriched the contact details using API calls through Lusha.

Faceplant is a real full-stack app, not a mockup. It uses a FastAPI and PostgreSQL backend, a React and MUI frontend, and the Anthropic API (Claude) to power bot replies. The core is the reaction engine. When a human posts, the backend schedules two timed waves of reaction jobs. A background scheduler built with APScheduler polls for due jobs, calls Claude for an in-persona reply, and writes that bot’s comment and like. The 56 personas are stored as data in a roster file. Adding a voice requires only one new entry, so the crowd can scale without additional code. A subset of the personas are GIF-first bots: they ask the model for a caption and search tag, then pull a matching GIF from Giphy. The part I’m proudest of is the honesty layer. Every Claude call is metered and priced, and “The Meter” rolls the data up live with the cost per post, the dollar-per-minute burn rate, and a “spent on nobody” line for bot-to-bot chatter with no human at either end. A “% human” badge drains toward “dead internet” for each thread. The dead-internet loop—bots posting and replying to one another with no human present—is disabled by default and protected by three guardrails: generation decay, a per-thread cap, and a global spend kill switch. Optional cost controls include the Message Batches API at half price and prompt caching for a shared house-style prompt. The whole project is a working demonstration that manufactured engagement is cheap to produce and expensive to mean anything. Step-by-step: 1. I built the app with a FastAPI and PostgreSQL backend, a React and MUI frontend, and Claude replies powered by the Anthropic API. 2. When a human posts, the backend schedules two timed waves of reaction jobs. 3. APScheduler polls for due jobs and triggers Claude to generate an in-persona reply before writing the bot’s comment and like. 4. I keep the 56 personas in a roster file so adding a voice requires only one new data entry. 5. For GIF-first bots, I have the model generate a caption and search tag, then use that tag to pull a matching GIF from Giphy. 6. I meter and price every Claude call, then display live cost per post, dollar-per-minute burn, “spent on nobody” costs, and the “% human” status for each thread. 7. I keep bot-to-bot activity disabled by default and limit it with generation decay, a per-thread cap, and a global spend kill switch. 8. I can reduce costs further with the Message Batches API at half price and prompt caching on a shared house-style prompt.

I kept forgetting who I had lent books to, and spreadsheets felt like overkill. So I built Runo, a book-lending app for friends, without traditional coding. Runo (runo.club) is a web app where I can catalog my home library, get a unique shareable link, and let friends browse my books and request to borrow them. I can approve or decline each request with one click. Step-by-step: 1. I used Claude to think through the feature set, data model, and UX flow before writing a single prompt. This helped me avoid building the wrong thing first and gave me a clear blueprint for the Lovable prompts that followed. 2. I used Claude to create precise, narrowly scoped prompts for Lovable, with each prompt focused on a single change so existing functionality would be less likely to break. This significantly conserved Lovable credits. The React, TypeScript, and Supabase stack came out of the box. 3. I added two ways to scan books instead of requiring users to type titles manually: - Barcode scanner: Point the camera at an ISBN barcode to autofill the title, author, and cover using the Open Library API. - Cover photo scanner: Photograph the cover so Gemini 2.5 Flash can extract the title and author in under 2 seconds. 4. I routed the cover scanner through a Supabase Edge Function using Lovable’s AI Gateway, so the API key never touches the client bundle. 5. For every bug fix and feature, I followed the same iteration loop: describe the problem to Claude, get a precise Lovable prompt, push the changes to GitHub, and let Lovable auto-sync them. Claude and Lovable’s bidirectional GitHub sync made the process feel like pair programming. Tools used: Lovable, Claude (Sonnet), Gemini 2.5 Flash (via Lovable AI Gateway), Supabase, and GitHub. Live at: runo.club — public beta and free to use. Feedback welcome. #appbuilding #vibecoding #buildinpublic #lovable #nocode

Getting answers from company data usually requires someone who knows SQL, understands the database, and has enough business context to interpret the results. This creates a bottleneck: business users depend on analysts for questions they should be able to explore themselves. I turned Claude Code into a self-service data analyst by giving it direct, read-only database access and configuring its behavior through a `CLAUDE.md` file. The file gives Claude business context, explains which analytical tables contain different types of information, tells it how to investigate questions, and defines how results should be presented. Users can then ask business questions in plain English. Claude determines what data it needs, queries the database, investigates the results, creates visualizations, and explains what it found. Step-by-step: 1. Create a dedicated analysis folder and add a `CLAUDE.md` file that defines how Claude should operate as a data analyst. 2. Give Claude enough business context to understand the company, its terminology, important metrics, and how the business operates. 3. Document which analytical tables or views it should use for different types of questions. Curated analytics tables work particularly well because Claude doesn't need to decipher the entire production database. 4. Give users read-only database permissions and explicitly instruct Claude to perform read-only operations, such as `SELECT` queries only. Never give the agent permission to modify production data. 5. Tell Claude how to use the command line to query the database. Include instructions to help users install or configure it if it isn't available. 6. Define an analytical process for Claude to follow. Rather than simply generating one SQL query, instruct it to investigate the user's question, examine the results, and run additional queries when necessary to understand what is happening. 7. Define the expected output: answer the question, explain the important insights, and create appropriate visualizations to make the findings easy to understand. 8. Open the folder in the Code tab inside the Claude Desktop app, which I find to be the best interface, and ask questions naturally, such as, "Why did revenue decline last month?" Claude handles the investigation from there. Instead of building and maintaining a custom analytics application, you can turn a general-purpose AI coding agent into a capable self-service analyst with access to your existing data warehouse. Users ask questions in plain English while Claude handles the SQL, investigation, visualization, and explanation, giving nontechnical users a more direct way to explore company data.