Can the agent commit code when using this skill?

No, the skill explicitly instructs the agent not to commit changes, leaving orchestration and version control management to the human user.

How does this skill handle different agent roles?

It separates agents into Test Writers and Implementers, restricting each role to specific file types to prevent overlap and ensure a clean TDD workflow.

Why are there specific shell rules included?

Claude's built-in Bash tools already capture output; these rules prevent agents from using unnecessary redirections that can cause command processing errors.

Does this skill work with any programming language?

Yes, it is designed to be language-agnostic and relies on your project's CLAUDE.md file for specific language contexts and test commands.

What is the purpose of the Agent Briefing skill?

It provides a standardized set of rules for AI agents to follow, ensuring they adhere to TDD practices, respect file boundaries, and maintain high code quality.

TDD Agent Briefing

Name: TDD Agent Briefing
Author: kelp

bykelp

•

セキュリティとテスト

Standardizes agent behavior and role-specific rules within a Test-Driven Development pipeline to ensure code quality and consistency.

The Agent Briefing skill provides a comprehensive operational framework for AI agents, establishing strict boundaries between test writing and implementation roles. It enforces high-quality standards by mandating minimum test coverage for all public functions, prohibits unauthorized modifications to build configurations, and optimizes terminal interactions by standardizing shell command execution. By integrating with project-specific context, it ensures agents remain aligned with local coding standards and directory structures, making it an essential tool for maintaining a robust TDD workflow.

主な機能

01Role-based permission enforcement for test writers and implementers

022 GitHub stars

03Strict file modification restrictions to protect build and config files

04Standardized shell command execution rules to prevent tool failures

05Seamless integration with CLAUDE.md for project-specific context

06Mandatory quality benchmarks with 100% public function test coverage

ユースケース

01Preventing AI agents from accidentally modifying critical build or infrastructure configuration files

02Standardizing AI behavior across a development team to ensure consistent code quality

03Coordinating multi-agent workflows where one agent writes tests and another implements features

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kelp/kelp-claude-plugins agent-briefing

For use in Claude.ai and ChatGPT

主な機能

01Role-based permission enforcement for test writers and implementers

022 GitHub stars

03Strict file modification restrictions to protect build and config files

04Standardized shell command execution rules to prevent tool failures

05Seamless integration with CLAUDE.md for project-specific context

06Mandatory quality benchmarks with 100% public function test coverage

ユースケース

01Preventing AI agents from accidentally modifying critical build or infrastructure configuration files

02Standardizing AI behavior across a development team to ensure consistent code quality

03Coordinating multi-agent workflows where one agent writes tests and another implements features