Skill Claude CodeCodex
A method for making AI-assisted coding measurable through a repeating cycle: assess the task, estimate it without seeing the result, implement it, review the outcome, and improve future estimates. It is available in Chinese and includes a set of web-development scoring criteria.
Skill Claude CodeCodex
A file-based handoff process for sharing software work between Codex and Claude Code. It moves a project through specification, planning, building, review, polishing, and committing.
Skill Claude CodeCodex
A workflow for one coding assistant to complete work through six controlled phases: specification, planning, building, review, polishing, and committing.
Skill Claude CodeCodex
A task-estimation workflow that scores coding tasks and records an advance prediction of the effort, risks, and likely outcome. It uses a separate agent for an isolated first assessment before review and storage.
Skill Claude CodeCodex
A workflow for revising the scoring model, or rubric, used to evaluate coding tasks. It requires enough completed examples, rescoring those examples, checking whether their order stays consistent, and getting an independent review.
Skill Claude CodeCodex
A setup guide for starting Vibe Code, a project evaluation tool, in either a new folder or an existing code project. It checks the project, asks setup questions, creates initial files, and can import its history.
Skill Claude CodeCodex
A code-review learning workflow that studies a high-quality open-source project and records its conventions in a benchmark file. The benchmark acts as a reference for judging future code.
Skill Claude CodeCodex
A schema migration tool for updating the version and structure of a .vibe-state.json file, a file that stores Vibe-related state. It follows the required version steps and checks that the updated file is valid.
Skill Claude CodeCodex
A codebase-profiling workflow that records a project's structure, complexity hotspots, file-change patterns, test coverage, and dependencies.
Skill Claude CodeCodex
A task-prioritization workflow that reads project state, available tasks, and previous assessments to suggest what to work on next. It separates lower-risk tasks from experiments intended to test an assumption.
Skill Claude CodeCodex
A Chinese-language review workflow for comparing predicted and actual results after a coding task is finished. It records the differences and updates project observations.
Skill Claude CodeCodex
INTERNAL sub-agent for blind 5-dim rubric scoring. NOT user-facing — called via Task tool by vibe-assess/vibe-bump. Receives ONLY taskpath + rubricpath. Hard refuses .vibe-state.json, predictions/, retros/. Outputs strict JSON: 5 dims × {score 0-5, confidence, reason}.
Skill Claude CodeCodex
A lightweight Chinese-language scoring workflow for quickly rating a coding task across several dimensions. It prints the result without creating prediction files or changing project state.
Skill Claude CodeCodex
A codebase scanner that finds possible work by checking TODO comments, files changed often, missing tests, outdated dependencies, and unusually large files.
Skill Claude CodeCodex
A read-only status dashboard for Vibe Code projects. It reads project state and handoff files, which record where work stands between planning, coding, review, and completion.
Skill Claude CodeCodex
A project trend scanner that checks software dependencies and related project information for recent changes. Dependencies are external packages that the project relies on.