arc-iac-mcp-expertlisted

Browse, search, scaffold, compare, and security-scan the 56+ SourceFuse ARC Terraform modules — works standalone via a bundled script, no MCP server required. Use this skill whenever the user wants ARC Terraform modules, asks "which ARC module should I use", wants to scaffold/generate Terraform from an ARC module (e.g. arc-eks, arc-db, arc-network, arc-vpc, arc-s3), compare two ARC modules, look up a module's inputs/outputs/resources/versions, find which modules create a given AWS resource, or tfsec-scan HCL before a PR — even if they don't name a tool explicitly.
urbanlotusai/arc-iac-skill · ★ 0 · DevOps & Infrastructure · score 72

Install: claude install-skill urbanlotusai/arc-iac-skill

# ARC IaC Expert Gives Claude direct access to SourceFuse's [ARC](https://registry.terraform.io/namespaces/sourcefuse) Terraform module catalog (56+ modules, live from the Terraform Registry): browse, search, inspect, scaffold production Terraform, compare modules, and run tfsec security scans. **This skill is self-contained.** It bundles [scripts/arc.py](scripts/arc.py) — a dependency-free Python 3.8+ CLI (stdlib only) that talks straight to the public Terraform Registry API and a hosted tfsec service. No MCP server, no `pip install`, no API key. It ports the logic of the `arc-iac` MCP server (`urbanlotusai/arc-iac-mcp`) so it works on any machine. > If the `arc-iac` MCP server *is* connected (tools named `arc_list_modules`, `arc_scaffold`, …), you may use those tools instead — they're equivalent. But never assume they're present; default to the bundled script. --- ## Running the script ```bash python3 <skill-dir>/scripts/arc.py <command> [args] ``` All commands print JSON (or, for scaffold/scan, formatted text) to stdout. Errors go to stderr with a non-zero exit. Pipe JSON through `python3 -c` or `jq` when you only need part of it. | Command | What it does | |---|---| | `list [--sort downloads\|name]` | All modules: name, description, downloads, version | | `search <query>` | Keyword search over names + descriptions | | `get <name> [--readme]` | Full detail: inputs, outputs, resources, versions (README omitted unless `--readme`) | | `inputs <name>` | Input variables