Skip to content

ThothCTL

AI-Powered Infrastructure Lifecycle CLI — scan, generate, review, and govern your IaC from a single tool.

  • Security Scanning


    5 integrated scanners (Checkov, Trivy, KICS, OPA, TF-compliance) with unified reports and enforcement.

    Scan your code

  • Agent Companion Development


    Build IaC with Kiro or Claude as your agent — ThothCTL's 26 MCP tools give AI agents hands to scan, generate, review, and deploy on your behalf.

    AI-DLC Workflow

  • Intent-to-IaC Generation


    Natural language → governed Terraform via CLI or agent. Org rules enforced at generation time, scaffold-grounded, self-correcting.

    Generate IaC

  • Inventory & SBOM


    CycloneDX 1.6 compliant SBOM, technical debt scoring, and dependency staleness tracking.

    Track dependencies

  • Cost & Drift


    AWS cost projections, ITIL v4 blast radius, and AI-powered drift detection.

    Analyze costs

  • DevSecOps Workflows


    Composable YAML pipelines with 7 SDLC phases. Hard/soft enforcement. CI/CD ready.

    Run workflows


30 Seconds to First Result

pip install thothctl
thothctl scan iac -t checkov     # scan existing IaC

Or use the guided setup:

thothctl quickstart              # interactive onboarding

How It Works

%%{init: {'theme':'base', 'themeVariables': { 'primaryColor':'#e3f2fd','primaryTextColor':'#1565c0','primaryBorderColor':'#1976d2','lineColor':'#42a5f5','secondaryColor':'#fff3e0','tertiaryColor':'#f3e5f5','fontSize':'14px'}}}%%
graph LR
    agent["<b>Agent / CLI</b><br/><small>Kiro · Claude · CLI</small>"]:::node0
    generate["<b>Generate</b><br/><small>Intent → IaC</small>"]:::node1
    scan["<b>Scan</b><br/><small>Security check</small>"]:::node2
    review["<b>Review</b><br/><small>AI analysis</small>"]:::node3
    deploy["<b>Deploy</b><br/><small>Blast radius</small>"]:::node4
    monitor["<b>Monitor</b><br/><small>Drift & cost</small>"]:::node5

    agent --> generate --> scan --> review --> deploy --> monitor
    monitor -.->|"feedback"| agent

    classDef node0 fill:#6200ea,stroke:#4a148c,stroke-width:2px,color:#fff
    classDef node1 fill:#7c4dff,stroke:#6200ea,stroke-width:2px,color:#fff
    classDef node2 fill:#ff9800,stroke:#e65100,stroke-width:2px,color:#fff
    classDef node3 fill:#e91e63,stroke:#880e4f,stroke-width:2px,color:#fff
    classDef node4 fill:#4caf50,stroke:#2e7d32,stroke-width:2px,color:#fff
    classDef node5 fill:#00bcd4,stroke:#006064,stroke-width:2px,color:#fff

Two interaction modes — same governed lifecycle:

  • Agent companion (Kiro, Claude) — AI agent drives ThothCTL via MCP. You chat, it executes. AI-DLC Guide
  • Direct CLI — You run commands directly. Scriptable, CI/CD-native. How It Works

Choose Your Path

Use Kiro CLI or Claude as your companion — the agent calls ThothCTL via MCP:

# Start the MCP server (agents connect to this)
thothctl mcp server

# In Kiro CLI:
kiro-cli chat --agent thoth
# → "Scan my terraform for security issues"
# → "Generate a VPC with 3 private subnets"
# → "What's the cost estimate for this stack?"

The agent has access to 26 MCP tools: scan, generate, review, inventory, cost analysis, drift detection, and more — all governed by your org rules.

Full AI-DLC Guide

cd my-terraform-project
thothctl scan iac -t checkov -t trivy     # security audit
thothctl inventory iac --check-versions   # dependency check
thothctl check iac -type cost-analysis    # cost estimate
thothctl quickstart                       # guided setup
# or:
thothctl init space -s my-space -vcs github
thothctl init project -p my-infra -s my-space
thothctl generate iac \
  -i "VPC with 3 private subnets and NAT gateway" \
  -p ollama --apply
# GitHub Actions
- run: thothctl scan iac -t checkov -t trivy --enforcement hard
- run: thothctl inventory iac --check-versions
- run: thothctl ai-review analyze -p bedrock

What Makes ThothCTL Different

ThothCTL Other tools
Scope Full lifecycle (generate → scan → review → deploy → monitor) Single-purpose
Agent-native 26 MCP tools — agents (Kiro, Claude) use ThothCTL as their hands No agent integration
AI Multi-agent review + intent-to-IaC + agent companion workflow None or basic
Cost Open-source, local-first, offline capable (Ollama) SaaS, vendor-locked
Output Standard Terraform/HCL — no runtime lock-in Proprietary formats
Governance OPA/Rego + org rules enforced at generation time Post-hoc scanning only

Supported Platforms

Framework Scan Generate Inventory Check Document
Terraform
OpenTofu
Terragrunt
CDK v2

Requirements: Python 3.10+ · Linux, macOS, or Windows (WSL)


Next Steps