Claude Standard Dev Team skill
Twelve subagents plus a standard-team playbook that carry a Claude Code project from PRD to release, with QA checks at each step.
- 106 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- Agents write and modify project code, including deployment files.
- A long multi-role loop burns many tokens.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/claude-standard-dev-teamDetects 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.
- Agents write and modify project code, including deployment files.
- A long multi-role loop burns many tokens.
This is third-party code. Review the repository files before installing.
What it does
The repository holds twelve agent roles and one `standard-team` skill: a product manager, architect, UI designer, database optimizer, backend and frontend developers, DevOps, a QA evidence collector, a security engineer, a code reviewer, a final acceptance checker and a technical writer. The main session runs the skill, which drives eleven phases and calls the roles through Task. Work is contract-driven: PRD, API contract and DB schema come first, then implementation. Each endpoint is verified separately and a failure sends the task back. After release an iteration mode assesses affected phases and runs evidence-based regression.
Who it is for. Solo developers and small teams building mid-size apps with a frontend, backend, database and deployment in Claude Code.
Good fit when
- There is an app idea and a PRD, API contract and DB schema are needed before code.
- Each finished endpoint has to be verified independently and returned on failure.
- Deployments broke on API paths and hard-coding them must be banned.
- A live project is being extended and only the affected phases should rerun.
Not a fit when
- The task is small and fits a single edit.
- Browser checks are required: the evidence collector works without browser automation.
- Tokens are tight: a twelve-role loop spends many requests.
Example request
Use the standard dev team: take a booking service description, produce the PRD and API contract, and stop at the first human checkpoint.Limitations
Twelve roles and eleven phases spend plenty of tokens and suit mid-size apps rather than small edits. The author notes the playbook was moved from an agent to a skill on 22 May 2026 because nested subagents did not work then; newer Claude Code versions change this and a return is being tested. The README and docs are in Chinese. Token use grows with the number of roles, so try it on a small project first.
How to disable. Delete the copied files from `~/.claude/agents/` and the `~/.claude/skills/standard-team/` folder.
FAQ
Do I install both agents and the skill?
Yes. Agents alone will not react to the standard-team phrase without the `standard-team` skill.
How many phases are there?
Eleven. Checkpoints for your confirmation follow the PRD and the API contract.
Can I take only some roles?
Yes. Copy, for example, `product-manager.md` and `software-architect.md` into `~/.claude/agents/`.
What does the final reviewer do?
`reality-checker` asks for rework by default and passes a release only on convincing evidence.
What happens after launch?
Iteration mode assesses affected phases, changes the contract and reruns only those.
Discussion
No questions or reviews yet.
Related
SkillsA set of Claude Code utilities: background skills, expert agents, slash commands and prompt templates for development
Claude Code kit with eight phase agents, 21 skills and hooks that keep the agent inside set limits.
A Claude Code working configuration: a planner with per-milestone review, sub-agents and skills for analysis, refactoring and docs sync
Six commands for Claude Code, Codex and Cursor: the main session holds context while fresh agents review the problem and diff.
A plugin that turns Claude Code into a 16 agent team: documentation first planning, TDD and 9 pre-merge quality gates.
A Claude Code plugin for development with separated roles: spec, review, TDD, QA and verification by different agents
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| Claude Standard Dev Team (this card) | no | 106 | |
| Awesome Claude Code Subagents | no | 26,506 | a large collection of standalone roles; standard-team links twelve roles in a playbook with checkpoints |
| agency-agents-zh | no | 22,040 | a set of role prompts in Russian and Chinese; here roles work in phases with contracts and QA |
README in short
Role architecture, eleven phases and version history. `INSTALL.md` has step-by-step installation and `WORKFLOW.md` describes the playbook.