forge

🥈Silver

The Vibe Coding++™ platform enables orchestration of multiple AI agents for collaborative coding. Developers can experiment with isolated attempts and ship code they understand. It features multi-agent kanban with MCP integration, streamlining project management and task automation.

78130Updated 1mo ago

Intermediate30min to implementautomation

Saves ~240 min per use

Quick InstallView Source

git clone https://github.com/automagik-dev/forge.git

Works with:

Claude

Overview

About This Skill

How to Use

[{"step":"Define the project scope and agents. Open Vibe Coding++™ and create a new project. Use the kanban template to assign agent roles (e.g., 'Senior Dev Agent' for implementation, 'Security Agent' for reviews) to stages like 'Design,' 'Implementation,' and 'Testing.' Tip: Start with 3-5 agents to avoid complexity; add more as the project scales.","action":"Navigate to 'Projects' > 'New Project' > 'Kanban Template' in Vibe Coding++™."},{"step":"Integrate MCP tools for automation. Configure the DevOps Agent to connect with your CI/CD pipeline (e.g., GitHub Actions, Jenkins) via MCP. Enable the QA Agent to pull test cases from your test management tool (e.g., TestRail) and push results to your bug tracker (e.g., Jira). Tip: Use MCP’s 'auto-trigger' feature to run tests on code commits.","action":"Go to 'Settings' > 'MCP Integrations' and map tools to agents (e.g., 'QA Agent' → 'TestRail' for test cases, 'DevOps Agent' → 'GitHub Actions' for deployments)."},{"step":"Run isolated experiments. Use the 'Sandbox' mode to test refactoring strategies without affecting the main branch. For example, if refactoring a payment service, clone the repo to a sandbox, let the Senior Dev Agent propose changes, and validate with the Security Agent before merging. Tip: Use the 'Diff Viewer' to compare sandbox changes against the original code.","action":"Select a repository in Vibe Coding++™ > Click 'Sandbox' > 'New Experiment' > 'Start Refactor.'"},{"step":"Monitor progress and iterate. Review the kanban board daily to track agent outputs. Use the 'Metrics Dashboard' (MCP-integrated) to monitor KPIs like code coverage, security score, and deployment frequency. Tip: Set up Slack/email alerts for critical stage completions (e.g., 'Security Agent flags a vulnerability').","action":"Open the 'Dashboard' tab in Vibe Coding++™ to view real-time metrics. Configure alerts in 'Settings' > 'Notifications.'"},{"step":"Document and ship. Once all stages are complete, export the kanban board, test reports, and security findings to your project management tool (e.g., Notion, Confluence). Use the 'Export' feature to generate a markdown report or JSON file for stakeholder review. Tip: Include a 'Lessons Learned' section in the report to refine future multi-agent workflows.","action":"Click 'Export' > Select format (Markdown/JSON) > Choose destination (e.g., Confluence, GitHub Wiki)."}]

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/automagik-dev/forge

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Use the Vibe Coding++™ platform to [TASK] with multi-agent collaboration. Set up a kanban board in the platform with [AGENTS] assigned to [STAGES]. Integrate with [MCP_TOOLS] for automation where needed. Document the process in [FORMAT] for review. Example: 'Refactor the legacy authentication system in the e-commerce app, assigning the security agent to validate changes and the testing agent to automate regression tests, then export the kanban board as a markdown table.'

Example Output

### Multi-Agent Collaboration Report: Legacy Authentication System Refactor

**Project Overview:**
The team used Vibe Coding++™ to refactor a monolithic authentication system in an e-commerce app (12,000+ daily users) over a 3-week sprint. The goal was to modularize the codebase, improve security, and automate testing while maintaining 100% uptime during the transition.

**Kanban Board (Multi-Agent Workflow):**
| Stage               | Agent Role               | Assigned Agent       | Status       | Outputs                                                                 |
|---------------------|--------------------------|----------------------|--------------|-------------------------------------------------------------------------|
| **Backlog**         | Requirements Analysis    | Product Owner Agent  | Completed    | 15 user stories prioritized; 3 critical security gaps identified        |
| **Design**          | Architecture Planning    | System Design Agent  | Completed    | Modular auth service blueprint (5 microservices); threat model updated   |
| **Implementation**  | Code Refactoring         | Senior Dev Agent     | In Progress  | 80% of legacy auth logic migrated to new `auth-service` (4500 LOC)      |
| **Security Review** | Penetration Testing      | Security Agent       | Completed    | 12 vulnerabilities patched (CVSS scores reduced from 7.8 to 3.2 avg)     |
| **Testing**         | Automated Regression     | QA Agent             | In Progress  | 234/250 test cases passing; 16 new edge cases added                      |
| **Deployment**      | CI/CD Pipeline           | DevOps Agent         | Pending      | Blue-green deployment strategy configured; rollback plan documented     |

**Key Achievements:**
1. **Security:** Reduced attack surface by 65% through modularization and input validation.
2. **Performance:** Authentication latency dropped from 420ms to 180ms (57% improvement).
3. **Automation:** 92% of regression tests now run in CI/CD via the QA Agent’s MCP-integrated scripts.
4. **Documentation:** Auto-generated API docs and threat model exported to Confluence.

**Challenges & Mitigations:**
- **Challenge:** Legacy session management broke during migration.
  **Solution:** Implemented a dual-write pattern with temporary backward compatibility.
- **Challenge:** Security Agent flagged a race condition in token refresh logic.
  **Solution:** Added distributed locking via Redis; tests now run in parallel safely.

**Next Steps:**
- Complete testing and deploy `auth-service-v2` to staging.
- Monitor error rates and performance for 48 hours post-deployment.
- Schedule a retrospective with the Product Owner Agent to refine the workflow.

**Metrics Dashboard (MCP-Integrated):**
- Code Coverage: 94% (target: 90%)
- Security Score: 9.2/10 (target: 8.5/10)
- Deployment Frequency: 3x/week (target: 2x/week)
- Mean Time to Recovery (MTTR): 12 minutes (target: 30 minutes)