Agentic-config is a project-agnostic, composable configuration system for AI-assisted development workflows. It serves as a single source of truth for agentic tools like Claude Code, Antigravity, Codex CLI, and Gemini CLI. Benefits developers by streamlining configuration management and improving workflow consistency.
git clone https://github.com/WaterplanAI/agentic-config.gitagentic-config is a centralized, project-agnostic configuration system designed for AI-assisted development workflows. It provides 42 composable skills organized across seven plugins: workflow orchestration, git automation, QA testing, utilities, meta-prompting, security guardrails, and audit logging. The system works natively with Claude Code plugins and pi packages, allowing skills to invoke other skills for compounding automation. Teams benefit from reproducible, tagged releases pinned in version control, while developers get streamlined configuration management without project-specific modifications.
1. **Initialize the Config**: Create a new `agentic-config.yml` file in your project root. Use the template above as a starting point, replacing `[PROJECT_NAME]`, `[TOOLS]`, `[LANGUAGE]`, and `[ENVIRONMENT]` with your project-specific details. 2. **Customize Workflows**: Define workflows for common tasks (e.g., testing, linting, deployment) under the `tools` section. Include commands, outputs, and dependencies for each step. 3. **Set Team Roles**: Under the `team` section, specify team size and roles. Assign shared workflows to roles to ensure consistency. For example, all backend developers should use the same testing workflow. 4. **Validate the Config**: Use the `agentic-config validate` command (if available) or a YAML validator to check for syntax errors and required fields. Reference the schema if provided. 5. **Integrate with Tools**: Configure your AI tools (e.g., Claude Code, Antigravity) to use this config file. For example, set the `AGENTIC_CONFIG_PATH` environment variable to point to your `agentic-config.yml`. **Tips for Better Results:** - Use comments in the config file to explain complex workflows or decisions. - Test workflows locally before sharing them with the team to ensure they work as expected. - Keep the config file version-controlled and update it as your project evolves.
Orchestrate multi-agent spec workflows with pimux-backed task delegation
Automate git operations, PR creation, and release management across repositories
Run E2E testing and browser automation in QA pipelines
Enforce security guardrails for credentials, write scope, and supply chain risks
No install command available. Check the GitHub repository for manual installation instructions.
git clone https://github.com/WaterplanAI/agentic-configCopy the install command above and run it in your terminal.
Launch Claude Code, Cursor, or your preferred AI coding agent.
Use the prompt template or examples below to test the skill.
Adapt the skill to your specific use case and workflow.
Generate an agentic-config configuration file for [PROJECT_NAME] that integrates [TOOLS] (e.g., Claude Code, Antigravity, Codex CLI). Include settings for [LANGUAGE], [ENVIRONMENT], and [ACTIONS] (e.g., testing, linting, deployment). Structure the config to support modular workflows and ensure consistency across [TEAM_SIZE] developers. Use YAML format.
```yaml
# agentic-config.yml for Project Phoenix
project:
name: phoenix-api
version: 1.2.0
language: Python 3.11
environment: production
tools:
claude_code:
max_tokens: 4096
temperature: 0.7
tools: ["file_operations", "shell", "web_search"]
sandbox: true
antigravity:
workflows:
- name: "test-suite"
steps:
- command: "pytest tests/ --cov=src --cov-report=xml"
output: "coverage.xml"
- command: "flake8 src/ tests/"
output: "linting-report.txt"
- name: "deployment"
steps:
- command: "docker build -t phoenix-api:${{GIT_COMMIT_SHA}} ."
output: "docker-build.log"
- command: "kubectl apply -f k8s/deployment.yaml"
output: "k8s-deploy.log"
codex_cli:
default_model: "codex-4"
hooks:
pre-commit:
- command: "poetry run black src/ tests/"
output: "black-format.log"
- command: "poetry run isort src/ tests/"
output: "isort-format.log"
team:
size: 8
roles:
backend: ["alice", "bob"]
frontend: ["charlie", "diana"]
devops: ["eve"]
shared_workflows:
- "test-suite"
- "linting"
validation:
required_fields: ["project.name", "tools.claude_code", "team.size"]
schema: "https://raw.githubusercontent.com/agentic-tools/agentic-config/main/schema.json"
```
### Key Features Demonstrated:
1. **Modular Workflows**: The `test-suite` and `deployment` workflows can be reused across the team, ensuring consistency in testing and deployment processes.
2. **Tool Integration**: Configures Claude Code for interactive development, Antigravity for automated workflows, and Codex CLI for pre-commit hooks.
3. **Team Collaboration**: Defines roles and shared workflows, making it easy for 8 developers to collaborate without conflicting configurations.
4. **Validation**: Includes a schema reference to ensure the configuration adheres to the latest standards.
This configuration ensures that every developer on the Phoenix team uses the same tools, commands, and workflows, reducing setup time and minimizing errors during development and deployment.Google's multimodal AI model and assistant
Fast, flexible flat-file CMS for modern websites
AI assistant built for thoughtful, nuanced conversation
AI-powered design tools for product teams
Identity and access management for agents
Design smarter sales workflows visually
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan