Terraform plans,
in plain English

Post your plan JSON. Get back a human-readable summary, risk flags, reviewer checklist, and PR-ready markdown. One API call. Any CI pipeline.

$ curl api.planplain.dev/analyze \
  -d @tfplan.json
Summary
Adding 2 resources, modifying 1,
destroying 1 (RDS instance).
Risk Flags
● HIGH: RDS instance being destroyed
● MED: Security group rule change
● LOW: New S3 bucket (standard)
Reviewer Check
☐ Confirm RDS backup exists
☐ Verify SG port range intended
Three lines in your pipeline
1

Plan

Run terraform plan and export JSON, like you already do.

terraform show -json tfplan > plan.json
2

Analyze

POST the JSON to PlanPlain. Get structured analysis in milliseconds.

curl api.planplain.dev/analyze -d @plan.json
3

Review

Drop the markdown into your PR comment. Reviewers know exactly what to check.

gh pr comment $PR --body "$RESPONSE"
Everything a reviewer needs, nothing they don't
📝

Plain-English Summary

What's actually changing, stated clearly. No parsing HCL diffs in your head.

"This plan creates a new production VPC
with 3 subnets, attaches an internet
gateway, and updates the route table
for public access."
⚠️

Risk Flags

Destructive changes, security implications, and cost impact flagged automatically.

HIGH - aws_db_instance.prod will be destroyed
MED - IAM policy scope widened to *
LOW - New CloudWatch log group

Reviewer Checklist

Context-aware checks tailored to the specific changes. Not generic boilerplate.

☐ Confirm database backup before destroy
☐ Verify IAM wildcard is intentional
☐ Check subnet CIDR for overlap
📄

PR-Ready Markdown

Copy-paste into GitHub, GitLab, or Bitbucket. Or automate it in CI.

## Terraform Plan Summary
**3 to add** | **1 to change** | **1 to destroy**

### Risk Assessment
| Level | Resource | Reason |
|-------|----------|--------|

Your team shouldn't need Terraform expertise to review infrastructure PRs

PlanPlain turns opaque plan output into clear, actionable reviews. Works with Terraform and OpenTofu. Fits any CI pipeline. No platform lock-in.