terrashark

Diagnose Terraform/OpenTofu failure modes before generating code changes.

314|24|Updated Feb 23, 2026

Search Tags:#terraform #infrastructure-as-code #opentofu #failure-modes #llm-safety #auditable-output

Ready to run

npx skills add https://github.com/LukasNiessen/terrashark --skill terrashark

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: terrashark
Source: https://github.com/LukasNiessen/terrashark/tree/main
Command: npx skills add https://github.com/LukasNiessen/terrashark --skill terrashark

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Terraform/OpenTofu generation can produce syntactically valid but operationally dangerous changes due to hallucinations. TerraShark fixes this by forcing a diagnostic workflow that identifies failure modes before producing code.

Core Features & Use Cases

Failure-mode-first diagnosis to surface risk prior to code changes
Load only the relevant references from a granular library (19 files) to minimize token usage
Migration playbooks and structured output contracts to enable auditable changes
Works across Terraform and OpenTofu, including module refactors, migrations, and CI/CD pipelines

Quick Start

Ask TerraShark to diagnose failure modes in a Terraform/OpenTofu task and generate a safe remediation plan with an auditable output contract.

Frequently Asked Questions about terrashark

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I prevent hallucinations when generating Terraform or OpenTofu infrastructure as code?▼

To prevent hallucinations in Terraform or OpenTofu, apply a failure-mode-first diagnosis workflow to surface operational risks before generating code, ensuring changes are auditable and safe.

What is a failure-mode-first workflow for infrastructure as code?▼

A failure-mode-first workflow for infrastructure as code diagnoses potential operational failure modes prior to code generation or refactoring, enforcing an explicit output contract to produce safe, auditable changes.

Can I use this diagnostic workflow for both Terraform and OpenTofu migrations?▼

Yes, this workflow supports both Terraform and OpenTofu, applying failure-mode diagnosis during module refactors, migrations, and CI/CD pipeline building to mitigate risks across both tools.

How do I review Terraform code for operational risks before applying changes?▼

Review Terraform code for operational risks by running a failure-mode diagnosis to identify dangerous changes, then load targeted references to generate a structured remediation plan with an auditable output contract.

Does this approach work for building CI/CD pipelines with infrastructure as code?▼

Yes, it works when building delivery and testing pipelines by diagnosing failure modes in the infrastructure as code tasks, enforcing safe generation practices, and providing structured migration playbooks for auditable changes.