The infrastructure control plane for workflows and agents
Every infrastructure change, from people or agents, in one place
Plan, review and apply Terraform, OpenTofu and Terragrunt with agents that review changes and open fixes, in our cloud or your own cluster, with your own models.
No credit card required for our starter tier
Stop rubber-stamping infrastructure PRs
A 400-line plan gets a quick LGTM because nobody has time to read it, and the pull requests nobody can review just sit there. InfraDots agents read every change with its plan, so your reviewers start from the findings.
Reviews that read the plan
Every pull request is reviewed together with its plan: what gets created, changed, replaced or destroyed, checked against the checklists for the resources it touches.
Fixes, not just comments
Findings can become a pull request: the implement agent writes the fix, and follow-up requests add to the same pull request.
Reviews to your standards
Your team's reviewer guidance is applied to every review, so the agent flags what your team cares about.
Workflows and agents, in our cloud or yours
One workflow for people and agents
Workspaces for Terraform, OpenTofu and Terragrunt, plan and apply runs with approvals, connected to GitHub, GitLab or Bitbucket.
Agents that review
The review agent reads the change with its plan, checks it against risk-based checklists, comments on the lines that matter and can request changes.
Agents that fix
Ask from the web app, Slack, the CLI or your IDE: the implement agent writes the change and opens a pull request, and follow-ups land on the same one.
Your cluster, your models
Run the agents and the Terraform executor in your own Kubernetes cluster with our public Helm chart, on your Anthropic, Bedrock, Vertex AI or Azure AI Foundry account.
Run it on your laptop
idp agent review --offline reviews your branch before you push. Free, with your own model key and no account.
Know what's changing
Drift checks, usage insights, variables and secrets, teams and permissions, and a Terraform provider and CLI to manage it all as code.
No credit card required for our starter tier
Use Cases
Built for teams who've outgrown manual infrastructure
Platform teams, startups and regulated engineering orgs use InfraDots to review, implement and apply infrastructure changes, with every change reviewed and on record.
For Platform & DevOps Teams
One Place for Every Infrastructure Change, from People or Agents
Approvals
On plan and apply runs
RBAC
Teams and permissions per workspace
Slack, web, CLI
Where changes can be requested
For Startups & Scale-Ups
Migrate Your Cloud Infrastructure to IaC in 4 Weeks
4 weeks
From console chaos to fully codified infra
Reviewed
Every change before it reaches production
6 months free
If we miss the 4-week deadline
For Engineering Teams
Agents that review and fix your Terraform PRs
Every PR
Reviewed together with its plan
Line comments
On the code, with changes requested when needed
One PR
Findings and follow-up fixes stay together
For Regulated Teams
Regulated teams: agents in your own cluster
Your cluster
Agent runner and Terraform executor
Your model account
Anthropic, Bedrock, Vertex AI or Azure AI Foundry
HTTPS out only
Runners connect outbound
Need help getting there? Our engineers migrate you in 4 weeks.
We review your cloud setup, plan the migration and codify it with your team, with every change reviewed before it hits production. 4-week guarantee: if we don't deliver, your first 6 months are free.
Frequently Asked Questions
What is InfraDots?
InfraDots is the infrastructure control plane for workflows and agents: one place for every infrastructure change, whether a person or an agent makes it. You plan, review and apply Terraform, OpenTofu and Terragrunt through workspaces with approvals, while agents review your changes and open pull requests with fixes.
What do the agents do?
The review agent reviews every pull request together with its plan, checks it against risk-based checklists, comments on the lines that matter and can request changes. The implement agent writes a change you ask for, from the web app, Slack, the CLI or your IDE, and opens a pull request for your review; follow-up requests add to the same pull request. Agents look up the newest provider and module versions in the public registries, and follow your team’s reviewer guidance.
What is a workspace in InfraDots?
A workspace is the execution unit that binds your Infrastructure as Code to a Version Control System(VCS) where your code resides. The workspace is tied to a branch of the VCS and a folder if there is a folder structure (monorepo).
What is concurrency in InfraDots?
This is simply the ruleset that limits how many workspace executions and agent reviews can happen at the same time
What is an interaction in InfraDots?
Interactions are the units we use to measure how many times our agents were invoked and took action: a review or an implementation. Local runs with `idp agent --offline` never reach InfraDots, so they are not interactions.
Is InfraDots cloud-agnostic?
Yes, InfraDots is completely cloud-agnostic. Whether you're using AWS, Azure, Google Cloud, or any other cloud provider, our platform can help you manage your infrastructure regardless of where it's hosted.
How does InfraDots improve developer productivity?
InfraDots provides self-service capabilities that empower your development teams to provision and manage infrastructure without waiting for operations teams. This reduces handoffs, eliminates waiting periods, and allows developers to focus on building features instead of managing infrastructure.
What support does InfraDots provide for infrastructure planning?
We offer human collaboration with our expert team who can help you map out your infrastructure roadmap and pipeline. Our team works with you to understand your specific needs and create a tailored infrastructure strategy.
What infrastructure frameworks does InfraDots support?
InfraDots supports Terraform, OpenTofu and Terragrunt.
Can I run InfraDots in my own cluster, with my own model?
Yes, on the Growth and Enterprise plans. Self-hosted runners run the agents and the Terraform executor in your own Kubernetes cluster from our public Helm chart, and connect out to InfraDots over HTTPS only. The agents use your own model account: Anthropic, AWS Bedrock, Google Vertex AI or Azure AI Foundry.
Can I try the agent without an account?
Yes. Install the idp CLI and run `idp agent review --offline` in your repository to review your branch before you push, or `idp agent implement --offline "add versioning to the logs bucket"` to get a change on a new local branch. It runs on your machine with your own model key, is free, and needs no InfraDots account.