Think inside your AI world.

Terraform through Unl

The Terraform MCP server will hand back a plan's full JSON, resource change by resource change. Whether one of those changes is a destroy you didn't sign up for is a rule your own team defines - a plan that runs isn't the same as a plan that's safe.

Terraform through Unl reads a run's plan JSON against the no-surprise-destroy rule you have ratified, so 'is this plan actually safe to apply?' comes back blocked with the condition it hasn't met.

What Terraform holds

Query the Terraform MCP server about a plan or workspace and it returns:

  • provider and module documentation pulled from the registry
  • organisation, project, and workspace listings in HCP Terraform
  • runs in a workspace, and detailed information about a specific run
  • the structured JSON output, logs, and details of a run's plan
  • Sentinel policies by topic, and details for a specific stack

What the naked read gives you

The Terraform MCP server pulls live provider and module documentation straight from the registry, searches Sentinel policies, lists organisations, projects, and workspaces in HCP Terraform, and retrieves the full JSON output, logs, and details of any run's plan. It reports exactly which resources a plan will create, change, or destroy, without judging whether any of that destruction is one you'd want to sign off on.

The frame judges the data it is given; it does not verify the source’s accuracy.

What changes when Terraform is read measured

A platform engineer at a healthcare SaaS company ratified an apply rule in Unl: no Terraform plan applies to the production workspace if its plan JSON shows a delete action against any resource tagged data-tier, because a plan once slipped through that silently destroyed and recreated a production database subnet group during business hours.

'Is this plan actually safe to apply?' returns, measured: not yet - the plan JSON read shows a delete action queued against aws_db_subnet_group.primary, a resource tagged data-tier, exactly the change the rule blocks. Terraform supplied the plan's resource changes; your ratified rule supplied the block.

And back again

Reworking the change - moving the resource instead of destroying and recreating it - is what unblocks this. The revised plan JSON shows no delete against the data-tier resource, and Terraform through Unl reads that straight into the existing page.

The answer comes back measured against what you already decided, and why.

The lane is live and open to this tool today: one box, paste anything. If it speaks MCP, Unl can reach it. Readings arrive unprompted, the data beside the criterion; Unl is a courier, not a warehouse, and keeps only your keys and the frame.

Questions people ask

What would clear this plan to apply?

Reworking the change so it no longer deletes a data-tier resource - a move or in-place update instead of a destroy-and-recreate. The same ratified rule reads the revised plan JSON and clears it.

How do I connect Terraform to Claude?

Run HashiCorp's official Terraform MCP server locally via Docker or binary, or point Claude at a remote deployment, authenticating against the public registry and, optionally, an HCP Terraform token for workspace and run access - Unl reads through that same server.

Does Unl write to my infrastructure?

No. Unl calls the server's search, get, and list tools - plan JSON, workspace details, run status - and triggers no apply. Any apply, you or your pipeline run it.

What this is

Think inside your AI world — you stay in command

Unlimitless (Unl to friends) holds what you've settled, reads what your tools are showing, and catches what's changed out in the world — and hands your AI whatever bears on the work, the moment it's needed, without you asking. The right thing, in front of the model, unprompted, with you in command of the call. So you keep moving toward what you set out to build, on top of everything you've already decided.

It plugs into Claude, Claude Code, ChatGPT and Cursor as an MCP connector. Quick to connect, in a couple of steps.

Unlimitless is open now to invited Alpha. Apply for the Beta waitlist to come in ahead of the full launch:

Alpha is invite-only · free at launch.