67 real things people did with Claude Code, with the source and a prompt to try it yourself.
About 15k pages of discovery stood between a solo practice and a summary judgment motion. Claude ran OCR, built embeddings and search scripts, then found evidence for each assertion, with traces to check.
Result Trusted because it left traces to follow
Markdown files should just open, so they asked Claude Code for a viewer. It opens instantly: faster, they say, than any other macOS app, and faster than Bitwarden's Chrome extension shows its fill menu.
Result Came out of the same codebase as their JSON viewer
Migrating a password reset flow, Claude Code and Opus produced code on par with a mid level developer. It needed close supervision throughout, and cost roughly 2 months of Pro subscription for about half a day's work.
Result A bit slower overall than doing it by hand
A side project needed a launch video, but not an agency's bill. Claude Code built one from the codebase plus their feedback. It took a lot of iteration over 2-3 days, all on subscription token usage.
Result No screen recording needed, the source was already there
A URL shortener side project ran on AWS. Claude Code wrote the GitHub Actions workflows for the move and fixed each failure as it came up. The writer reports zero downtime.
Result Estimated hosting: about $44 a month, down from $120
A 7-app monorepo needed an 8 GB memory setting and 25 seconds to typecheck. Claude Code made the changes for the move to oxlint, oxfmt and tsgo, and the team validated each step.
Result No more 8 GB memory setting in CI
WhatsApp has no native way to schedule messages, and the writer didn't want third-party apps in their chats. Claude Code built a scheduler using a headless browser, and a recurring message now goes out every morning.
Result Later added a Telegram bot, no extra AI tokens spent
Possible business expenses were buried in a stack of credit card statements. Claude Code went through every statement, flagged the charges that might need reimbursing and put them in an Excel sheet.
Wanting a backup of their own LinkedIn posts before the platform's bot-blocking caught up, the writer had Claude Code drive a headless browser through their feed and copy the posts to their personal blog.
Result Switched to logging in manually when LI caught on
A roughly 50-page tax return, run through three coding agents side by side. Codex was accurate first time in 12 minutes, Antigravity missed a value until prompted, and Claude Code stumbled overnight.
Result Filling the forms by hand would have been faster anyway
Before a conference with 40 people they had never met, LinkedIn research normally takes an evening. Claude Code researched the attendees in parallel via LinkedIn and Chrome MCP, producing a full briefing in 5 minutes.
Result Also books flights and hotels in the company travel system
A recorded rapid chess game came with a voice memo of what the player was thinking move by move. Claude drove Stockfish to question each thought and built a narrated video explaining the mistakes.
Result Answered the exact question asked aloud at move eight
A thumbnail research task used to take three full days of manual work. Claude Code built a tool analysing 776 videos across 57 channels to find what worked, one of four tools a non-coder built in three weeks.
Result Also rebuilt their website in under an hour
One message from a phone through Claude Code's Dispatch: boot World of Warcraft and log in to a specific character. Claude Code opened Battle.net, launched the game, waited out the login and picked the character.
Result Asked to draw in Paint, it wrote Python instead
Turning plan files into Claude Code workflows took 5-6 hours each. A script that converts them deterministically, and flags what can move out of the LLM, cut that to 20-30 minutes at about 20% of the tokens.
Result Also removed a 10-15 minute workflow construction wait
Writing a weekly newsletter kept losing out to more urgent work. A Claude Code skill now reads new blog posts, drafts the email in the writer's own voice, and creates the broadcast in Kit once it's approved.
Result Some months got zero newsletters; now one goes out weekly
An inherited app on Azure, an error, and a hunt that could easily take 1-2 days by hand. With CLI access to Azure and the repo, Claude Code ran the config and network checks and found the root cause in about 45 minutes.
Running Claude Code in YOLO mode risked projects leaking tokens, so a container setup isolated each project's permissions. Typing a hotkey and a feature name, like text-to-speech streaming, got it done in five minutes.
Result From nothing on screen to words at every pause
Claude proposed the changes, opened a Jira issue with small subtasks, and shipped them as reviewable PRs with tests. A coworker had guessed two weeks; Claude and ChatGPT guessed 6 to 8 weeks.
Result Shipped as reviewable PRs with tests
Detailed Markdown specs go to OpenClaw, which spawns and tests Claude Code instances for hours until the app works as described. "Is the code any good? Probably not." They check before trusting it.
Result Sometimes builds for 2+ hours unattended
Reviewing every turn kills the value. Working from a spec and a to-do list, Claude Code runs for weeks on a compiler and a strategy game while they step in now and then.
Result A spec and a to-do list keep it on track
They never had the time or energy for the backlog. Claude Code writes small, deterministic programs that do the work, and most of the scripts they'd wanted for years now exist.
Result Programs do the work, not the AI each time
Ask about the page you're reading and the chat already has its text. Claude Code built the Streamlit app in a single file, correctly, on the first attempt.
Result Page text added to every question automatically
They already run local reviews with Codex and Claude Code. Assigned as a reviewer in GitHub, Copilot finishes in a couple of minutes and raises valid issues they'd never found.
Result Review done within a couple of minutes
Learning enough to fix it would have taken them a week. Claude Code diagnosed it the same afternoon and built a small tool that fixes it.
They worked with Claude Code like an engineering team: requirements, design specs, test plans, step by step for about 3 months. The app scores QR codes before the phone acts on them.
Result Started as just for them; now polished
They replaced their bookkeeper with Claude connected to QuickBooks. Asked to check last year's return, it found the CPA hadn't depreciated goodwill from an acquisition, and the CPA agreed.
Result A clean rerun missed the same deduction
Days or months after highlighting a line, it appears on the display on their dresser. Claude Code wrote the backends for the e-ink screens, showing a random quote with its book cover, 100% local.
Result Each screen now shows a different quote
Recipes from photos, nutrition from the USDA, a weekly plan and shopping list. Their meal data showed they already get most vitamins, and 2 weeks of symptom logs point to tomatoes as a first suspect.
Result First suspect for their symptoms: tomatoes
The Android apps for their old Brother printer were terrible and ad-ridden. With Claude Code in the browser on their phone, they built their own print page, one of dozens of little family apps.
Result Costs basically nothing to run
A read-only bank feed, synced email and beancount, all wired up by Claude Code so the accounts stay current. At year end it advised through a complicated return; an IRS review found one small error.
Result Books kept current automatically, all year
By hand it would easily have taken a day. Claude Code cleared them while they watched TV, and has since added a feature to a Rust editor they don't even know.
Years of Oh My Zsh for just 3-4 features, and a slow shell to show for it. Claude Code rebuilt those features into a single .zshrc, and the shell got much faster. Their other configs moved in minutes.
Result Moved to Lazy.nvim and Ghostty in minutes too
Always-on agents are overkill for a schedule. Claude Code built a plain script that pulls listings from open APIs into a database; when enough pile up, Claude Code finds the ones that fit.
Result Simple, predictable, AI only where it helps
Onboarding onto a new project means a stream of unstructured thoughts. They dictate them, and Claude Code tags and links everything in Obsidian. Weeks later, "What did I do yesterday?" just works.
Result "Which project uses Java 21?" now answerable
Give it detailed tasks, agree the approach, then let it iterate for hours. Their 20 years of experience steps in only for obscure bugs, and sometimes they play with the dog while it works.
Result Now working like a tech lead, not a coder
Which bike races are on TV today, and where? The data is messy and the sites break every week. Each day Claude reads the unmatched-race report and proposes fixes, so one person can keep it running.
Result Legal TV listings for every race, every day
They needed site and app blocks at set times that are hard to remove. They built their own, about 40% with Aider and the rest with Claude Code. It beats every commercial option they found.
Result Beats every commercial Linux blocker they tried
Alarms and reminders are easy to ignore. Claude built a Twilio script that calls before important appointments, reads out the message, and won't stop until they confirm. Set up in plain English.
Result Calls get through where alarms get ignored
They go on hour-long walks and just talk. A voice memo app they vibe coded sends the transcript to Claude Code, which turns it into structured notes. Some bugs, but it works.
Result Nearly 20K lines, vibe coded
Micromanaging five, 10 or 20 agent sessions doesn't scale. Their OpenClaw agent, Patch, runs the Claude Code instances in a shared tmux session and reports back in one Telegram chat.
Result One Telegram chat instead of 20 sessions
The recovered files were malformed. Claude Code helped work out that each was a metadata header, a thumbnail and the real image stuck together, then wrote a script that split them, metadata intact.
Result Way faster than doing it by hand
An undocumented legacy codebase mixing two languages, a SCADA platform and proprietary hardware logic. With little help, Claude Code worked out most of it, with only a couple of misunderstandings.
Result Ready for the refactor the team had planned
The secret is all in the spec. They pressure-test a structured one in Kiro for hours or days, then hand it to Kiro CLI and Claude Code in tmux. Usually the agents finish without stopping.
Result A one-line TDD rule lifted every test's quality
Thousands of words of planning on their phone over the holidays, then Claude Code built it. It forgot one key feature until asked, and their son's reaction was one of the best looks they've had.
Result Built with a test suite and a clear to-do list
An agency's day is morning briefs, meeting prep and client action items. They built their own OpenClaw-style assistant with Claude Code, plugged it into WhatsApp, and now run it all from phone or laptop.
Result Next up: family travel and childcare
They add analysis to sim racing lap guides programmatically. Claude Code drives DaVinci Resolve three ways: timeline files, Lua scripts and computer use. Computer use burns tokens, but it works.
Result Works, though computer use is token-hungry
They wanted real practice to learn SQL. Claude Code built 12 projects that get harder step by step, each with tests that check the answers. A few lessons needed fixing; overall it worked fantastically.
Result Worked fantastically, with a few fixes
Some of those bugs had beaten a 30-year veteran for over a week. Claude Code ran a /loop around the clock on a prototype compiler, fixing over 500 crashing test cases while they glanced in now and then.
Result Including bugs that beat a 30-year veteran
Surveys, supplier risk, emissions reports, modern slavery checks: the data gathering never ends. With Claude Code they built tools that now do about 70% of it. Plus a game of Frogger, for breaks.
Result Plus a game of Frogger, for breaks
Flakes, Sway, Home Manager, secrets, even vim, all asked for in prose. Claude Code set up the old ThinkPad, and when an update blacked out the screen, it found the open bug and rolled Gnome back.
Result Zero regrets, even after a black-screen bug
The game doesn't show all your cats at once, and you can have 100+. Within a day Claude Code built a PyQt6 tracker; two months on, it has 600 users and Chinese, Russian and Polish versions.
Result Translated into Chinese, Russian and Polish
Linux config files are where switching gets painful. Claude helped them pick Fedora, answered every setup question and had Claude Code write the scripts they'd never have bothered with.
Result Scripts they'd never have bothered to write
Authenticate, pick the cluster, find the namespace, run kubectl, every single time. Claude Code built a tool that remembers it all: "a huge time saver" they'd never have had time to build.
Result A huge time saver they'd never have built
Instead of plugins, the agent writes down improvement ideas, edits its own Rust source and recompiles itself. It runs on DeepSeek 4 Flash on two local GPUs at about 138 tokens a second.
Result About 138 tokens a second on local GPUs
They needed tooling to spawn a Go proxy and run integration tests, using only Perl's standard library. Claude Code wrote most of it, and it beat their usual shell scripts for stability.
Result More stable than their usual shell scripts
They wanted notes their way, so they built the app with Claude Code. Barely 10 hours in, it does almost everything they asked for, plus extras. The App Store rejected it, though.
Result Almost everything they asked for, plus extras
A small publisher couldn't afford in-house developers for its tedious ops and finance processes. A former engineer in a business role automated them with Claude Code; what's left is mostly QA.
Result Real shipped code, mostly QA left
Hooks firing at once on the same event clashed: two wrote the same state file and corrupted it. Running each event's hooks in turn from cached input fixed it, for about 200ms per prompt.
Result About 200ms overhead per prompt
A scanner flags CVEs; the real question is which ones can actually be reached. For every one Wiz flags, Claude Code runs a reachability analysis, for about $200 a month in usage.
Result Real risk checked for about $200 a month
When the time-tracking and invoicing tool they relied on hiked its price, they had Claude Opus recreate it and load in 13 years of their own data. Read-only works; editing is still in progress.
Result Read-only version working; editing in progress
A 4-year-old React anatomy app needed to reach more of the world. Claude Code internationalized it for 45 major languages in one session, covering ~85-90% of the world's population.
Unicode characters were rendering wrong in PDFs. Claude Code raced through several approaches in an afternoon to find each one's limits, ending in a custom font fallback. Then a day of polish.
The auth prototype came together fast, until the last step: it sent the user ID as a plain string next to the token. Caught in review and fixed, it still beat starting from scratch.
Result Caught in review, and still faster than scratch
A full .Net website and a React Native app, built entirely with Claude Code. It works, but the code is "definitely overengineered", the bugs keep coming, and the lesson was to commit early and often.
Result Lesson learned: commit early and often
Optimizing a Taichi kernel had each agent looping on its own. Putting Claude Code and Codex in one discussion to combine their ideas cracked it, saving several hours.
Vermont town meeting reports locked the budget away in PDFs. A volunteer had Claude Code and Gemini pull the numbers out and build a comparison site, deployed in under an hour. Rough, but it works.
Result Deployed in under an hour