Codex use cases: 23 real examples
Agentic Feats lists 23 sourced examples of Codex in real work; 14 link the artifact; most common job type: coding (9). Newest first.
- Feats
- 23
- Shown
- 1461%
- Industries
- 5
- Updated
- Oct 13
Four features finished in about 2 hours instead of 3-5 days each, and the daily view took 40 minutes.
Parallel feature shipping in Codex
Keith Rumjahn used the OpenAI Codex desktop app with git worktrees and parallel threads to build four features for their Mac app at once.
Codex caught real errors, including a wrong Schedule L field mapping, and the author found their own 2025 mistakes.
S-Corp taxes with two agents
Claude Code wrote Python scripts that compute the author's S-Corp tax forms from QuickBooks and bank data, while Codex reviewed the code.
The agent handles all kinds of tasks for their company, and the author reports no memory problems so far.
Codex agent as company employee
The author built an autonomous agent from Codex, ChatGPT Plus, systemd, and UNIX email and user accounts, and uses it like an employee.
The associate confirmed the refund and the author got $15 back within about five minutes, while the author watched it happen.
Amazon subscription refund
Codex running GPT-5.5 cancelled an Amazon subscription, opened the live chat with an associate, and argued for an immediate refund.
Codex returned a complete, classified list of missed email threads where Claude missed many; it also deployed a vacation site researched from hotel availability.
Codex email audit and trip site
Codex, driven by two authors, audited a month of missed emails, built a vacation planning site from a phone, and shopped for a dishwasher.
Codex returned a ranked comparison of three Belmond properties with availability and nightly rate ranges for a sample window, plus a more detailed Word doc.
Client notes to Belmond shortlist
Codex Desktop read a client notes Word doc, checked Belmond's live booking site, and built a three-option comparison matching the client's past trip preferences.
The author cut time on literature triage and experiment setup, and says related-work sections now need less citation double-checking. No numbers are given.
Research workflow automation with Codex
Codex runs a weekly arXiv search, triages new papers, feeds a Zotero library, and turns hypotheses into cheap experiment tests for an AI safety researcher.
Codex sorted 240 dinner recipes down to 62 family-fit picks, rated them, and produced an index, a spreadsheet, and a Google Drive folder.
Comms tasks with Codex
The author used OpenAI's Codex to build a PR coverage dashboard, sort a messy work folder, and filter a recipe site for picky kids.
The Codex agent resolved the issue in 40 minutes, and the ops agent confirmed the fix in Slack.
Slack issue triage to Codex fix
An internal ops agent triaged a Slack issue, passed it to a Codex agent, and then verified the fix back in Slack.
From one prompt it created a repo, wrote the app and tests, got CI green, provisioned Postgres and deployed behind HTTPS with no further messages.
Self-hosted agentic software factory
The author built a sandboxed home server stack where Hermes with Codex takes one prompt and runs the whole software lifecycle, from repo to deployment.
Claude gave failed builds for a week and used up the 20x Max weekly limit. Sol had worked 20+ hours without success.
VNC client build attempts
The author asked Claude, then ChatGPT 5.6 Sol, to build a bare-bones VNC viewer for Windows 10 that avoids dropped keystrokes.
Codex traced the kills to the Linux OOM killer on a 2 GB droplet with no swap, and listed fixes without changing anything.
Find why a process dies
Codex CLI, run on the author's Mac, connected to their Ubuntu server over SSH and checked kernel logs to find why Claude Code kept dying.
Get five like these every week
The week's strongest feats, each with its receipt, plus a short roundup of AI news. Free.
You're on the list.
A free weekly newsletter. See what's in it
About 1,570 entries posted in Claude Code alone, and Codex caught a VAT return error days before filing.
AI bookkeeping for Swedish companies
Claude Code and Codex kept the books for several Swedish limited companies over ten weeks, handling receipts, VAT, bank matching and posting.
The author uses Codex daily for reports, editable designs, video editing and this newsletter, which still needed plenty of corrections.
Substack stats report in Codex
Codex used Computer Use to open the author's Substack dashboard and pull figures for every published post into a newsletter report.
Each model pass added about 1.5x to 2.0x speedup, reaching roughly 7.5x to 32x over baseline; one implementation beat xgboost on speed.
Faster Rust ML code via benchmarks
Coding agents wrote Rust ML algorithms like UMAP from scratch, then iteratively optimized them against criterion benchmarks.
Codex built a full EDM track and a piano duet from scratch in minutes in Ableton, shown in two demo videos with the prompts listed.
Ableton Live MCP for Codex
The author built an MCP server so Codex can control Ableton Live, including running Python inside Live.
All three produced a playable multiplayer game with a map list and saved rooms; Opus won on visuals, Gemini on backend, Codex landed in between.
Three agents build multiplayer Counter Strike
The Instant team gave Codex, Claude Opus 4.5 and Gemini 3 Pro the same prompts to build a 3D multiplayer Counter Strike clone.
Claude handled analytics, pricing, staffing and simple stall placement well, but failed at pathways, large coasters, and vertical terrain. The mod took about 40 hours.
Claude Code plays RollerCoaster Tycoon
Ramp Labs forked OpenRCT2, built a kubectl-style CLI, and let Claude Code manage a theme park through it.
About 19k lines of Rust across 154 commits in three days. The engine passes 64 SQLLogicTest queries and 282 AI-written unit tests.
SQLite-like engine in Rust
Claude, Codex, and Gemini workers, two of each, built a SQLite-style database engine in Rust by claiming tasks through git lock files.
Built a working What's for Dinner app with pantry, garden, recipe and restaurant pages plus a suggestions engine, in under a week.
First personal dinner-planning app
A non-programmer used Codex to build a local web app that suggests dinners from pantry, garden, dietary rules, appliances and weather.
The author shipped a redesigned personal site, nearly right in one shot, after a few more iterations.
Personal site redesign with Codex
Codex CLI redesigned the author's personal site after reading a design prompt copied from Anthropic's frontend-design plugin into a local markdown file.
The author showered while Codex waited in the queue and negotiated. The refund was done when they returned.
Stolen package refund via Computer Use
Codex with Computer Use waited in Amazon's support queue, polling on a schedule, then pursued a refund for a stolen package.
A working prototype runs on the handset over mobile data: typed chat via DeepSeek, battery check, torch, calls and alarms, tested on the phone.
AI chat app on Nokia 110
Codex helped the author reverse-engineer a Nokia 110 4G's firmware and build a native AI chat app with a custom agent.
No feats match this filter. Show all.
Proof marks:Shown artifact linked·Numbers a specific number stated·Claim claim only