molyanov-ai-dev skill
molyanov-ai-dev is a set of skills and agents for Claude Code and Codex: feature planning, project knowledge, code review.
- 303 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Installation copies scripts and skills into agent home directories and can affect existing setup.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/molyanov-ai-devDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Installation copies scripts and skills into agent home directories and can affect existing setup.
This is third-party code. Review the repository files before installing.
What it does
The repository describes a methodology for building with agents and ships the skills for it: `project-initialization`, `user-spec-planning`, `code-writing`, `documentation-writing`, `code-reviewing` and more. The route depends on task size: a small change goes straight to an execution skill, a feature with unclear behavior goes through planning and approval. Durable project facts live in Project Knowledge under `.claude/skills/project-knowledge/`, while `CLAUDE.md` stays a short entry point. Claude files are the source, and Codex versions are generated by a sync script.
Who it is for. Developers running one project in Claude Code and Codex who want a shared workflow without a heavy process for every small task.
Good fit when
- A new repository needs setup and written project knowledge.
- A feature needs behavior agreed before any code is written.
- The same skill set must work in Claude Code and Codex.
- Finished changes are checked by separate reviewers.
Not a fit when
- One skill for a narrow task is enough, not a whole methodology.
- The team will not maintain project documentation.
- Installation must not touch global agent files.
Example request
Let's plan this feature: clarify behavior first, write a spec and wait for my approval.Limitations
Installation copies files into `~/.claude` and `~/.codex`, so compare existing configuration by hand and do not overwrite it wholesale. The Codex sync script creates and updates files, and conflicts need manual resolution. The methodology reflects its author and may not fit every process. One author maintains the repository, and the rules reflect their approach to development. A Russian README also exists. Technical skill instructions are in English, while user-facing messages follow the user's language.
How to disable. Remove the copied skills and agents from `~/.claude` and `~/.codex`, then drop the `~/.agents/skills` link if it was created for this set.
FAQ
Which runtimes are supported?
Claude Code and Codex. Claude files are the source, and Codex versions are generated by `sync-to-codex.sh`.
Does a small fix need the whole process?
No. A small well-defined change goes directly to the matching execution skill.
Where is project knowledge stored?
In `.claude/skills/project-knowledge/`. Its `SKILL.md` routes and loads only the context a task needs.
What is Project Knowledge?
A set of durable project facts in `.claude/skills/project-knowledge/`: purpose, architecture, code patterns, deployment.
How is feature planning started?
Tell the agent the feature must be planned. The `user-spec-planning` skill runs an interview and requires spec approval before implementation.
Discussion
No questions or reviews yet.
Related
SkillsTwo skills for Codex and Claude Code: prd keeps a living implementation checklist and review audits designs from the user's side
A Claude Code working configuration: a planner with per-milestone review, sub-agents and skills for analysis, refactoring and docs sync
14 skills that find the gaps in a task before the code: interviews, plans, assumption tests and proof that a bug fix works
Claude Code kit with eight phase agents, 21 skills and hooks that keep the agent inside set limits.
Tasks, docs, specs and decisions live as markdown files in your repository, and the agent reads them through the CLI and MCP
Over twenty skills for coding agents: MCP and hooks, plan mode, a bug-fix protocol, prompting and a bash safety hook
README in short
The README sets out the methodology, runtime pairs and the feature lifecycle: plan, validate, approve, implement, finalize.