equal-ai-toolkit

Equal AI Claude Skills

Production-grade Claude Code skills for SRE investigation and integration testing.

Available Tools

Tool	Type	Description
/investigate-alert	Command	SRE alert investigation with Datadog, CloudWatch, code analysis. Produces RCA reports.
integration-test-expert	Skill	Auto-triggers when discussing integration tests. Uses production data to inform coverage.

Prerequisites

Before installing, configure these MCP servers in Claude Code:

1. Datadog MCP (required)

   • Run /mcp → Add Datadog server with API credentials

2. Slack MCP (optional, for auto-fetching alerts)

   • Run /mcp → Add Slack server

Installation

/plugin marketplace add swap-inf/equal-ai-claude-skills
/plugin install equal-ai-toolkit@equal-ai-claude-skills

First-Run Setup

After installation, run:

/investigate-alert

You'll be asked three questions:

• AWS CLI profile name (e.g., ai-prod-read)
• Database username (e.g., equalreadonly)
• Database password

All Equal AI infrastructure values (ECS cluster, services, database hosts) are pre-configured.

Note: The config file contains credentials. Add .claude/config/toolkit-config.yaml to your .gitignore.

Configuration is saved to .claude/config/toolkit-config.yaml in your project.

Usage

SRE Investigation

# Investigate latest alert from Slack
/investigate-alert

# Investigate specific issue
/investigate-alert "5xx errors on api-gateway since 14:00"

# Investigate by Datadog monitor ID
/investigate-alert 12345678

Integration Test Expert

The skill triggers automatically when you mention:

• "create integration tests"
• "integration test suite"
• "test my API"
• "e2e test"

What's Included

SRE Investigation (/investigate-alert)

Structured 4-phase alert investigation:

1. Gather: Parse alert, build context
2. Investigate: Parallel subagents for infra + application
3. Validate: Build evidence chain, test alternatives
4. Report: Structured RCA with deep links

Uses:

• Datadog: Logs, metrics, APM traces, monitors
• AWS CloudWatch: ALB, ECS, RDS, Redis metrics
• Code Analysis: Trace through source code
• Slack: Fetch alerts from channels

Integration Test Expert

Builds comprehensive test suites by:

• Analyzing production traffic via Datadog
• Mapping API flows and dependencies
• Generating pytest tests with real DB/Redis fixtures

Configuration

After setup, config is at .claude/config/toolkit-config.yaml:

aws:
  profile: your-prod-profile
  region: ap-south-1

services:
  - name: user-service
    datadog_tag: user-service
    ecs_name: user-service-prod
    code_path: services/user-service/app/
    test_path: services/user-service/tests/integration/
    database: user_service_db

# ... more settings

See plugins/equal-ai-toolkit/config/example-config.yaml for full example.

Requirements

• Datadog MCP Server: For logs, metrics, traces
• AWS CLI: With read-only production profile
• Slack MCP Server: Optional, for fetching alerts

Updating

To get the latest version:

/plugin marketplace update equal-ai-claude-skills

License

MIT