openai/codex
17 skills7.8k installs2M starsGitHub
Install
npx skills add https://github.com/openai/codexSkills in this repo
1Babysit PrBabysit PR is a Codex skill that keeps watching an open GitHub pull request until it merges, closes, or hits a blocker that needs human help. It runs gh_pr_watch.py in snapshot or continuous watch mode, reads the actions list, diagnoses failed workflow jobs from GitHub Actions logs, and patches branch-related compile, test, lint, or snapshot failures while avoiding unrelated flaky or infrastructure errors. Review handling surfaces published human and Codex bot feedback, commits fixes with codex-prefixed messages, resolves eligible threads under a strict GitHub mutation policy, and never auto-replies to human reviewers without explicit confirmation. Flaky failures can be retried up to three times when the watcher recommends retry_failed_checks. The loop polls every minute while CI is red and keeps watching after green so late review comments are not missed. Terminal outcomes are PR closed, merge, or exhausted retries, permissions issues, or ambiguous reviewer requests.3kinstalls2Test TuiThe test-tui skill guides agents through interactive Codex TUI verification when validating Rust or CLI changes. Start the process interactively, always set RUST_LOG=trace for deep diagnostics, and pass -c log_dir=<temp_dir> so logs land in a known directory for post-run analysis. When sending test messages programmatically, write the text payload first and send Enter in a separate write; never combine text and Enter in one burst or the TUI may mis-parse input. Use the just codex target to launch, for example just codex -c log_dir=/tmp/codex-logs. Invoke when developers need to smoke-test Codex terminal UI behavior, reproduce TUI bugs with trace logs, or automate basic interactive sessions while preserving realistic keystroke timing. Pairs with Codex development workflows where terminal interaction must be exercised before shipping UI-facing changes. Start Codex TUI interactively with RUST_LOG=trace enabled.1.9kinstalls3Code ReviewRun a final code review on a pull request. Use subagents to review code using all code-review-* skills other than this orchestrator. One subagent per skill. Pass full skill path to subagents. Use xhigh reasoning.416installs4Code Review TestingTest authoring guidance. For agent changes prefer integration tests over unit tests. Integration tests are under `core/suite` and use `test_codex` to set up a test instance of codex.392installs5Codex Pr BodyUpdate the title and body of one or more pull requests. When this skill is invoked, the PR(s) to update may be specified explicitly, but in the common case, the PR(s) to update will be inferred from the branch / commit that the user is currently working on. For ordinary Git usage (i.e., not Sapling as discussed below), you may have to use a combination of `git branch` and `gh pr view <branch> --repo openai/codex --json number --jq '.number'` to determine the PR associated with the current branch / commit.276installs6Code Review ContextModel visible context. Codex maintains a context (history of messages) that is sent to the model in inference requests. The code-review-context skill documents workflows, constraints, and examples from SKILL.md for agent-assisted execution.275installs7Remote TestsTesting against remote executors in integration tests. Remote executor tests exercise the app-server/exec-server split to ensure that agent features work in both local and remote execution environments.264installs8Code Review Change SizeChange size guidance (800 lines). Unless the change is mechanical the total number of changed lines should not exceed 800 lines. For complex logic changes the size should be under 500 lines.263installs9Codex BugDiagnose GitHub bug reports in openai/codex. Use when given a GitHub issue URL from openai/codex and asked to decide next steps such as verifying against the repo, requesting more info, or explaining why it is not a bug; follow any additional user-provided instructions. Diagnose a Codex GitHub bug report and decide the next action: verify against sources, request more info, or explain why it is not a bug.252installs10Code Breaking ChangesGuide for testing Codex TUI interactively. You can start and use Codex TUI to verify changes. The test-tui skill documents workflows, constraints, and examples from SKILL.md for agent-assisted execution.251installs11Codex Issue DigestRun a GitHub issue digest for openai/codex by feature-area labels, all areas, and configurable time windows. Use when asked to summarize recent Codex bug reports or enhancement requests, especially for owner-specific labels such as tui, exec, app, or similar areas. Produce a headline-first, insight-oriented digest of `openai/codex` issues for the requested feature-area labels over the previous 24 hours by default. Honor a different duration when the user asks for one, for example "past week" or "48 hours". Default to a summary-only response; include details only when requested.224installs12Update V8 VersionUpdate Codex's pinned `v8` / `rusty_v8` versions, validate the release-candidate path, and investigate failed V8 canary or artifact builds. Use when asked to bump V8, update `rusty_v8` artifacts, prepare or validate a V8 release candidate, check `v8-canary`, or diagnose why a V8 version update no longer builds. 1. Read `third_party/v8/README.md` and follow its version-bump sequence. Treat that document as the release-process source of truth. 2. Inspect and update the concrete repo surfaces that carry the pin: - `codex-rs/Cargo.toml` - `codex-rs/Cargo.lock` - `MODULE.bazel` - `third_party/v8/BUILD.bazel` - `third_party/v8/README.md` - the matching `third_party/v8/rusty_v8_<version>.sha256` manifest when the remaining prebuilt inputs change 3. Keep the existing checksum helpers in the loop:117installs13Pushing Ci ChangesThe pushing-ci-changes skill explains how to push GitHub Actions workflow edits to the Codex repository when temporary approval is required. The repository blocks uploads to CI configuration under .github YAML paths unless the user holds a granted temporary role. Agents cannot request exemptions themselves and must direct the user to the internal workflow-approvals process before progress continues. When a CI push is imminent, the skill still recommends attempting the push to confirm whether the user account already has approval rather than assuming failure upfront. If git push rejects due to CI restrictions, share the workflow-approvals link, explain that only the user can request access, and ask them to return control after GitHub reflects the approval. The skill covers resolving push rejection and requesting upload exceptions without attempting bypasses. It is narrowly scoped to CI configuration changes rather than general repository contribution. Use when editing Codex GitHub Actions workflows or troubleshooting blocked pushes to CI files.78installs14Path TypesThe path-types skill documents path typing conventions in OpenAI Codex plugins and skills, helping authors reference files, commands, and resources with correct path schemas and validation rules.58installs15Plugin CreatorThe plugin-creator skill from OpenAI Codex guides authoring plugins for Codex including manifest structure, command definitions, skill bundling, and packaging workflows aligned with Codex plugin ecosystem conventions.30installs16Skill CreatorThe skill-creator skill guides creation of effective Codex skills that extend capabilities with specialized knowledge, workflows, and tool integrations. Core principles emphasize concise context because skills share the window with system prompts and history, appropriate degrees of freedom matched to task fragility, and protecting validation integrity when using subagents for realistic task evaluation. Anatomy requires SKILL.md with YAML name and description frontmatter plus optional bundled scripts, references, and assets folders. Description writing guidance stresses third-person trigger phrases, specific capabilities, and clear use-when boundaries since frontmatter alone drives skill selection. Workflow covers brainstorming, draft SKILL.md, resource bundling, validation on realistic prompts, and iterative refinement. Guardrails challenge every paragraph for token cost and avoid duplicating knowledge Codex already has. This skill provides guidance for creating effective skills. equipped with procedural knowledge that no model can fully possess.14installs17Imagegenimagegen skill documents Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, a. name: "imagegen" description: "Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code o8installs