Create photos, illustrations, product mockups, and other bitmap images, or edit existing images by replacing backgrounds, removing objects, and making transparent cutouts. Use reference images to guide style, composition, or visual variants.
Review code changes without modifying files, checking surrounding code, tests, and callers for defects introduced by the change. Receive actionable findings ranked by severity, with precise file locations and a brief assessment of remaining risks.
Find guidance on Codex settings, pricing, and OpenAI products and APIs using current official documentation. Get cited answers for setup, troubleshooting, model comparisons, migration planning, and prompt changes, with uncertainty stated when sources are inconclusive.
Browse available Codex skills and install them from a curated list or a specified GitHub location. Supports experimental skills, private repositories with credentials, and installing several skill folders in one operation.
Create or update Codex skills that capture task-specific guidance. Organize instructions in SKILL.md, add useful scripts, references, or assets when needed, and validate the structure while keeping the skill focused on its intended work.
Set up a realtime evaluation folder with a suitable shared harness, the framework that runs evaluations. Prepare prompts, tool definitions, realistic starter data, and a README, then validate the setup with evaluation and test runs.
Review invoices, expense receipts, and service contracts against accounts-payable policy before human approval. Recalculate invoice totals, flag missing documentation and risky contract terms, and produce a JSON report with findings and a review decision.
Review OpenAI Cookbook documentation and notebook prose for technical accuracy, readable wording, consistent terminology, and publication metadata. Make focused edits, preserve notebook code and outputs, and report unresolved issues with validation results.
Forward tasks from Claude Code's rescue subagent to Codex through a helper script. This internal skill handles model and effort options, fresh or resumed runs, and write access, then returns the helper's output unchanged.
Compose compact, structured prompts for coding, debugging, review, and research with Codex or GPT-5.4. This internal skill defines task scope, completion criteria, output format, verification steps, and evidence requirements using consistent XML tags.
Present results returned by a Codex helper while preserving findings, severity, file references, and uncertainty. This internal guidance also covers reporting edited files, failed runs, and setup requirements, with user confirmation before fixing review findings.
Capture a local static HTML page from the terminal using Playwright and Chromium. The supplied commands install the browser, set a 2048 × 1152 viewport, and save screenshots without starting a web server.
Compare local case files with insurer policies to prepare a concise prior authorization packet, covering approval requirements, referrals, missing information, and a checklist for staff reviewing the case and communicating with the patient.
Find and retrieve official OpenAI documentation to answer API integration questions with source evidence. Check endpoint definitions, parameters, and documented limits, and receive setup guidance when the documentation connection is unavailable.
Plan and, after explicit approval, run controlled checks of software behavior. Compare baseline and failure cases, repeat sensitive measurements, and report observed results with execution context, comparison conditions, and the status of temporary evidence files.
Run final checks on reviewed SDK changes, covering code formatting, linting for code issues, type checking, and tests. Confirm completion only after every required check succeeds, with wrappers for macOS, Linux, and Windows.
Compare a Python SDK release plan or candidate with the previous release to assess compatibility, versioning, and proven regressions. Produce a ship-or-block recommendation, documentation handoff, and key changes for minor releases.
Review completed code changes against the original requirements, choosing review depth by behavioral risk. Coordinate independent reviewers, resolve actionable findings, and verify that the delivered changes match the content actually reviewed.
Prepare a Python SDK release candidate in a separate working directory, check version metadata and public API compatibility, and produce one local commit with a review report for the release handoff.
Compare SDK implementation with English documentation to find missing features, incorrect defaults, and outdated behavior descriptions. Report evidence and suggested locations, or make requested updates while keeping navigation consistent and following the repository’s documentation checks.
Plan SDK behavior changes by defining required outcomes, compatibility boundaries, and unsupported cases. Reuse existing implementation paths, assess review feedback, and reconsider designs when repeated fixes introduce complexity beyond the original requirement.
Assess GitHub issues and pull requests by checking demonstrated user needs, existing alternatives, compatibility, and maintenance costs. Produce an evidence-based maintainer recommendation through code and document inspection, without running tests or examples.
Audit Python SDK logs, exceptions, and telemetry for sensitive data exposure. Trace values to their sources, fix confirmed leaks, and test that redaction hides sensitive information while preserving the caller's expected behavior.
Carry agreed implementation work through a separate Git worktree, an isolated working directory. After checks and review, it prepares one local commit, a clean branch, and a pull request draft for handoff without publishing them.
Is this your account? Link GitHub to bring it onto your creator page (Japanese)