Codex Stalls on a Merge Conflict or Resolves It the Wrong Way
Codex hits conflict markers and either stops, picks the wrong side, or hand-edits the markers into the file. Pre-rebase setup.sh and explicit AGENTS.md rules fix it.
#Codex
Articles tagged with #Codex
Codex Added a Package but the Lockfile Did Not Change
Codex edits package.json but never ran npm install, so package-lock.json diverges. How to enforce lockfile updates via setup.sh, CI, and AGENTS.md.
Codex Creates a New TypeScript Interface Duplicating One That Already Exists
Codex defines a fresh User or ApiResponse type when an identical one lives elsewhere. How to make the agent search first via AGENTS.md, indexes, and ts-morph.
Codex Amended or Rebased Git History You Did Not Want Touched
Codex ran git commit --amend or interactive rebase from an old README, rewriting shared history. How to lock the agent to forward-only commits.
Codex PR Description Says "Refactored Components" and Nothing Else
Codex emits a one-line PR body with no detail. How to force a structured description with before/after, why, and test plan via templates and AGENTS.md.
Codex Says Tests Passed but Actually Skipped the Failures
Codex reports green tests but the failing cases were filtered out, marked .skip, or bailed early. How to force honest test reporting before merge.
Codex Bails Out When the Patch Gets Too Large
Codex hits a context or token cap mid-task and stops with a partial diff. How to scope tasks smaller, split across PRs, or move to a larger-context model.
Codex Writes React.FC, componentWillMount, and Other Deprecated APIs
Codex outputs API patterns deprecated years ago because its training data is dated. How to lock it to modern APIs via AGENTS.md, ESLint deprecation rules, and reference docs.
Codex Committed Your Work to the Wrong Branch (or Straight to main)
Codex ran git commit on whatever branch was checked out — sometimes main, sometimes a leftover branch from the previous task. How to force a clean per-task branch.
Codex's Fix Passes Every Test but Breaks at Runtime
Codex's PR is green in CI — every test passes — but the app crashes in staging. Why agent fixes that target the test surface miss the runtime, and how to close the gap.
Claude vs Codex for PM Tasks: Which One Actually Saves Time
Two strong models, two different shapes of PM work. Here is the side-by-side on PRDs, JIRA grooming, and doc cleanup.
Codex Agent Output Conflicts With Prettier
Codex writes a file, Prettier reformats it on save, and now the diff is huge. Fix by running Prettier inside the agent loop and aligning agent style hints to your .prettierrc.
Codex Agent Goes Out-of-Context on Long Repos
On 500k+ line codebases Codex loses track halfway. Fix by scoping the working set, using directory summaries, and forcing the agent to read AGENTS.md before scanning.
Codex Agent Skips Files Containing Binary Data
Codex silently skips PNGs, PDFs, sqlite DBs, compiled artifacts. The fix is allowlisting text extensions, hashing binaries instead of reading, and giving Codex a binary-aware tool.
Codex Agent Spawns Too Many Redundant Tool Calls
Codex re-reads the same file 8 times and re-greps the same query 5 times. Fix by pre-feeding context, requiring a plan first, and capping reads via tool restrictions.
Codex Agent Stops Mid-Task Without Error
Codex halts halfway through a multi-step task with no visible failure — usually a context window, sandbox timeout, or stop-condition issue. Diagnose by checking the last tool call and turn budget.
Every Codex Keyboard Shortcut, Explained
A complete Codex keyboard shortcut reference for the terminal TUI, desktop app, input editing, session control, mode switching, background tasks, and failure cases.
Codex Skills and /skills, Explained: Built-in, Custom, Plugin, Team-Shared
A complete Codex skills guide: what a skill is, how /skills discovery works, how to invoke skills, how to write SKILL.md, and how built-in, plugin, and custom skills differ.
Codex Cannot Access a Private Repo
"unable to clone" — usually the Codex GitHub App isn't installed on the org, the repo isn't in its allowlist, or your OAuth token expired.
Codex Environment Setup Fails Before the Task Starts
Setup script bails before Codex can touch code — usually Node/Python version mismatch, private registry auth, or a slow install hitting the sandbox timeout.
Codex PR Too Large to Merge
50 files, 1500 lines, nobody can review it. Cap diff size at the prompt + AGENTS.md level; split refactors into shippable chunks.
Codex Beginner Guide: Sandboxed Cloud Tasks Without the Pitfalls (2026)
What Codex does, how it runs sandboxed cloud tasks, and when to use it. Setup, the spin-off workflow, and the mistakes that bite first-time users.
Codex for Code Review: Catch Bugs Before PR Review
Use Codex as a pre-review pair — catches issues before humans see them.
Use Codex to Review Your Sitemap
Use Codex to spot-check sitemap correctness, freshness, and coverage.
Codex Audit Report Is Too Broad to Act On
You asked Codex for an audit; got a 50-bullet report mixing typos with architecture. Re-scope by dimension, cap output size, force file:line.
Codex Cannot Finish a Patch Cleanly
Codex stops mid-patch — broken imports, half-converted types, untouched call sites. Bind "done" to a verifier, not a vibe.
Codex Duplicates Files Instead of Editing Existing Ones
Codex creates `utils.v2.ts` next to `utils.ts` instead of editing in place. Force in-place edits with explicit AGENTS.md rules + git as the safety net.
Codex Fails to Run or Interpret Build Results
Codex skips the build, misreads the output, or trusts a truncated tail. Use machine-readable verifiers + exit codes, not prose summaries.
Codex Fixes One Bug But Breaks Nearby Logic
The reported bug is gone, two adjacent features regressed. Cap blast radius, enumerate call sites, prefer caller-side guards over shared-util changes.
Codex Ignores Your Project Structure
New files land in `/src/` when your repo uses `/app/`; deps go to the root `package.json` in a monorepo; tests use a framework you don't have. Fix with AGENTS.md + canonical-example pointers.
Codex Makes Unsafe Assumptions
Codex confidently uses a column that doesn't exist, an env var with the wrong shape, or a function signature it never read. Force quote-before-write grounding.
Codex Misses Project-Specific Conventions
Codex uses `getUserById` when your codebase uses `findUserById`. Codify each convention as `rule + canonical example` in AGENTS.md.
Codex Patch Conflicts With Existing Code
"Patch hunk failed to apply" — file moved between Codex's read and write. Refresh state, normalize whitespace, or split the patch.
Codex Code Review Quality Feels Shallow
"Consider error handling" / "add tests" — bullets that work on any PR. Fix with specific review questions tied to file:line.
Codex-Generated Code Does Not Fit the Existing Style
Functionally correct but reads like a stranger wrote it — async/await mixed with .then(), wrong import ordering, comment style off. Anchor to a canonical file + enforce with lint.
Codex Test Suggestions Are Too Generic
"Test happy path and error path" — useless in 2026. Force tests bound to the function's actual signature, edge cases, and bug history.
Codex Output Doesn't Match Project Structure
Codex creates files in the wrong place or with wrong conventions — constrain with AGENTS.md and explicit refs.
Designing Prompts for Website Building (So AI Doesn't Drift)
Keep AI agents on-spec for multi-file website work with this CLAUDE.md template, 3-part prompt structure, and per-task constraint pattern.
Doing an SEO Review of Your Site with Codex
Use Codex (or any code-aware AI agent) for a technical SEO review with copy-paste prompts, dist/ inspection commands, and a structured-data validator.
Codex vs Claude Code: Which AI Coding Agent Should You Use?
OpenAI Codex and Anthropic Claude Code are both AI coding agents — but they live in different places and solve slightly different problems. Here's a clear comparison.