redc/SKILLS.md

Redc Skills - AI Operations Guide

Multi-cloud red team infrastructure automation tool built on Terraform.

Quick Reference

Config: ~/redc/config.yaml | Templates: ~/redc/redc-templates/ | Docs: https://github.com/wgpsec/redc

Configuration

providers:
  aws:
    AWS_ACCESS_KEY_ID: "KEY"
    AWS_SECRET_ACCESS_KEY: "SECRET"
    region: "us-east-1"
  aliyun:
    ALICLOUD_ACCESS_KEY: "KEY"
    ALICLOUD_SECRET_KEY: "SECRET"
    region: "cn-hangzhou"
  tencentcloud:
    TENCENTCLOUD_SECRET_ID: "ID"
    TENCENTCLOUD_SECRET_KEY: "KEY"
    region: "ap-guangzhou"

Env vars: Set AWS_ACCESS_KEY_ID, ALICLOUD_ACCESS_KEY, TENCENTCLOUD_SECRET_ID, etc. if config.yaml unavailable.

Global Flags

Flag	Default	Description
--config	~/redc/config.yaml	Config file path
-u, --user	system	User identifier
--project	default	Project name
--debug	false	Debug mode

Commands

Setup

redc init                           # Initialize templates
redc pull <image>                   # Download template (e.g., aliyun/ecs)
redc image ls                       # List local templates

Deploy

redc run <template> -n <name> -e key=val    # Plan & start (create infrastructure)
redc plan <template> -n <name>              # Plan only (preview without creating)
redc start <case-id>                        # Start case (create infrastructure)

Manage

redc ps                             # List all cases
redc status <case-id>               # Check status
redc stop <case-id>                 # Stop infrastructure
redc kill <case-id>                 # Stop & remove
redc rm <case-id>                   # Remove case

Execute

redc exec <case-id> <command>       # Run command
redc exec -t <case-id> bash         # Interactive shell
redc cp <src> <case-id>:<dest>      # Upload file
redc cp <case-id>:<src> <dest>      # Download file

Workflows

Quick Deploy:

redc pull aliyun/ecs && redc init
redc run aliyun/ecs -n myserver -e password=Pass123
# Returns case-id (e.g., 8a57078ee856)
redc exec 8a57078ee856 whoami

Controlled:

redc plan aws/ec2 -n staging
# Review the planned resources, then:
redc start <case-id>
redc cp deploy.sh <case-id>:/root/
redc exec <case-id> "bash /root/deploy.sh"

Cleanup:

redc stop <case-id> && redc rm <case-id>
# Or: redc kill <case-id>

Automation

import subprocess, re

def redc_run(template, name, env=None):
    cmd = ["redc", "run", template, "-n", name]
    if env:
        for k, v in env.items():
            cmd.extend(["-e", f"{k}={v}"])
    result = subprocess.run(cmd, capture_output=True, text=True, check=True)
    match = re.search(r'[a-f0-9]{12,64}', result.stdout)
    return match.group(0) if match else None

# Usage
case_id = redc_run("aliyun/ecs", "auto_deploy", {"password": "Secure123"})

Output Patterns

• Success: ✅ in output
• Error: ❌ in output
• Case ID: [a-f0-9]{64} (use first 12 chars)
• Status: running, stopped, created, error

Error Handling

Error	Solution
Config not found	Create ~/redc/config.yaml
Template not found	Run redc pull <template>
Case ID not found	Check redc ps
SSH timeout	Verify instance running, security groups
Init failed	Check network, configure Terraform mirror

JSON Schemas

Case:

{
  "id": "string[64]",
  "name": "string",
  "template": "string",
  "status": "running|stopped|created|error",
  "outputs": {"public_ip": "string"}
}

Config:

{
  "providers": {
    "aws": {"AWS_ACCESS_KEY_ID": "string", "region": "string"}
  }
}

Best Practices

• Use short IDs (first 12 chars)
• Assign meaningful names: <project>_<purpose>_<env>
• Always cleanup: redc stop → redc rm
• Use --debug for troubleshooting
• Never commit config.yaml to version control
• Monitor costs with redc ps

Resources

• Repo: https://github.com/wgpsec/redc
• Templates: https://github.com/wgpsec/redc-template
• Online: https://redc.wgpsec.org/

---

Version: 1.0.0 | License: Apache 2.0