Refresh: Best Infrastructure as Code (IaC) Tools for 2026 — tables, versions, quotes, FAQs by workprentice[bot] · Pull Request #20609 · pulumi/docs

workprentice · 2026-07-31T12:26:18Z

Summary

Refreshes /blog/infrastructure-as-code-tools/ — our #1 flagship cited asset in the Infrastructure as Code category (86 citations) and the page targeting our highest-value head keyword — for staleness and for a specific, measured gap on Perplexity.

Why: mention vs. citation gap on Perplexity

Pulling Profound visibility and citation data for the Infrastructure as Code category (2026-07-01 to 2026-07-30):

Pulumi's mention rate (visibility) is strong and roughly comparable across engines: ChatGPT 83.5%, Google AI Overviews 76.5%, Perplexity 72.8%.
But on Perplexity specifically, pulumi.com ranks only Do not link @pulumi/cloud in the AWS quickstart #7 among cited domains (130 citations, 3.07% share), behind spacelift.io (Switch to using typedoc for TypeScript documentation. #1, 288 citations), sentinelone (Update in the wake of renamepocalypse #2), talent500 (Improve our TypeScript documentation story #3), env0 (Quickstart for LM release #4), and apiiro (Set up monitoring and alerting for docs.pulumi.com #6).
Pulumi is Switch to using typedoc for TypeScript documentation. #1 domain overall in the category by total citations (422, 15.6% share), so the gap is not brand authority — it is that Perplexity is a citation-first engine that links to extractable, structured source data, and this page was under-supplying that structure (one table in 1,430 lines).

The fix in this PR is structural, not just cosmetic: more dense, self-contained reference tables, an original measured statistic, and named-source quotes — the concrete signals Perplexity and other answer engines link out to.

Competitor teardown

Audited the two pages currently outranking/out-citing us for this term:

spacelift.io ("16 Most Useful IaC Tools for 2026"): only 1 table (an end-of-article nav matrix), no answer-first opening, no FAQ section at all.
env0.com ("14 Best IaC Tools... Expert Picks"): similar single-table, no-FAQ structure.

Our existing FAQ section + faq_schema: true FAQPage markup is already an advantage neither competitor has. The clear gap to close was table density and per-tool pricing/licensing/version facts.

What changed

Table density: 1 → 4 tables. Upgraded the master comparison table (added License + Latest stable release columns) and added three new tables: pricing at a glance, license/governance/release-cadence, and a "which tool fits your situation" decision matrix. Each has a one-line lead-in sentence for extractability.
Answer-first opening. Replaced throat-clearing intro with a direct 40–60 word answer naming the shortlist, folding in a freshness marker and an original statistic.
Original statistic (measured 2026-07-31): Pulumi's own package registry lists 380 package entries spanning 308 unique packages (first-party providers, bridged Terraform providers, and community components) — framed as packages, not providers, since the registry mixes provider and non-provider entries.
Version/license refresh for all 10 tools using current GitHub release data (as of 2026-07-31): Pulumi v3.255.0, Terraform v1.15.8 (license corrected from "BSL 1.1" to the accurate "BUSL-1.1"), AWS CDK v2.262.2, Azure Bicep v0.46.1, Kubernetes v1.36.3, Crossplane v2.3.4, OpenTofu v1.12.5.
Staleness fixes: "150+ providers" → "170+ providers" (matches the sitewide convention used 67x elsewhere vs. 4x for "150+"); bumped example nginx:1.21 container image tags to nginx:1.27.
3 named, sourced expert quotes placed in context: Matt Stephenson (Starburst) on migration deployment speed, Yarin Miran (Wiz) on multi-cloud scale, Joe Duffy (Pulumi CEO) on AI-agent readiness.
3 new FAQ entries (pricing in 2026, which tools work best with AI coding agents, most popular tool by GitHub stars) — picked up automatically into the existing FAQPage schema since faq-entity.html auto-detects ### ...? headings under the FAQ section; no front-matter change needed for this.
Byline updated to "Pulumi Content Team" (existing pulumi-content-team.toml entity, already used on the sibling /blog/best-terraform-alternatives/ post); updated/lastmod bumped to 2026-07-31.

Schema notes (verified against layouts/partials/schema/, not assumed)

faq_schema: true was already correctly set and wired to graph-builder.html → faq-entity.html; the 3 new FAQ headings are auto-collected with no schema change required.
itemlist / itemlist_name front matter is unchanged and still correct (10-tool list is unchanged).
The card asked to verify comparison-table schema still validates on this page. Finding: comparison-table-entity.html is explicitly gated to URLs containing /comparisons/ and requires a literal | Feature | table header — this is a /blog/ URL, so comparison-table schema was never applicable here by design. Noting this so it isn't mistaken for a regression.
Did not add howto_schema: this page is a listicle, not a step-by-step tutorial, so that schema would not accurately describe the content.

Data sources

Profound (category: Infrastructure as Code) visibility and citations reports, 2026-07-01–2026-07-30.
GitHub REST API for current release versions/dates/licenses/stars (queried 2026-07-31).
Pulumi, HashiCorp, and Spacelift pricing pages plus corroborating third-party sources (queried 2026-07-31); one figure (Spacelift's listed annual price) was excluded from the page because its billing unit could not be confirmed — accuracy over density.
pulumi.com/registry/packages.md for the original package-count statistic.
Direct case-study and press-release sources for all quotes (linked inline).

🧠 This PR was created by workprentice on behalf of the Pulumi SEO/AEO content team — no get_me-equivalent tool was available in this session to resolve a specific requester's username.

…otes, FAQs - Add 3 new reference tables (pricing at a glance; license/governance/release cadence; decision matrix) to increase citable table density from 1 to 4 - Upgrade master comparison table with License and Latest stable release columns - Answer-first opening leading with the tool shortlist and a measured registry statistic (380 package entries / 308 unique packages, measured 2026-07-31) - Refresh version, license, and release-date facts for all 10 tools (Terraform BSL 1.1 -> BUSL-1.1; add current stable versions/dates for Pulumi, Terraform, AWS CDK, Bicep, Kubernetes, Crossplane, OpenTofu) - Fix stale 150+ providers references to 170+ (sitewide convention) and bump stale nginx:1.21 example image tags to nginx:1.27 - Add 3 named, sourced expert quotes (Starburst, Wiz, Pulumi CEO) tied to migration speed, multi-cloud scale, and AI-agent readiness - Add 3 new FAQ entries (pricing, AI-agent readiness, most popular tool by GitHub stars) picked up automatically by the existing FAQPage schema - Update byline to Pulumi Content Team; bump updated/lastmod to 2026-07-31

github-actions · 2026-07-31T12:37:38Z

Pre-merge Review — Last updated 2026-08-04T00:52:21Z

Tip

Summary: This PR refreshes the evergreen listicle content/blog/infrastructure-as-code-tools/index.md for 2026 — it rewrites the intro as a direct answer, adds three new comparison tables (pricing, license/governance/release cadence, "which tool fits your situation"), stamps a Latest stable release line onto each of the ten tool sections, adds three customer/executive pull-quotes, and appends three FAQ entries. Because the whole value of the refresh is currency, the wrongness that would hurt a reader most is a wrong version, date, price, or license in one of the new tables — a reader planning a 2026 tool decision will take those numbers at face value. Most of them hold up: Pulumi v3.255.0, Terraform v1.15.8, AWS CDK v2.262.2, Kubernetes v1.36.3, and OpenTofu v1.12.5 all match their upstream GitHub releases, as do the license, governance, and star-count columns and the Pulumi/HCP Terraform pricing figures. @CamSoper disputed the Azure Bicep version finding and was right — a fresh check of Azure/bicep's GitHub releases confirms v0.46.1, published 2026-07-30T23:35:23Z, non-prerelease, is genuinely the latest release; the original check's source evidently returned a stale snapshot. All three Bicep occurrences (L138, L176, L697) are conceded to ✅ Resolved. Two blockers remain: the Spacelift/env0 pricing sentence is wrong about those vendors publishing tiers, and the new AI-agents FAQ asserts that HCL lacks native unit testing, which stopped being true with Terraform 1.6. Passes run: external claim verification (78 claims across four specialists), cited-claim spot-checks, a frontmatter sweep, a temporal-trigger sweep, prose linting, and the editorial-balance pass. Cross-sibling reads and code-example execution did not apply.

Review confidence:

Dimension	Level	Notes
mechanics	MEDIUM	Frontmatter adds `lastmod` alongside `updated`, against the repo convention; the byline was reassigned away from two named authors.
facts	MEDIUM	48 of 78 claims verified against upstream sources; the Azure Bicep version/date pair, initially flagged as wrong, was confirmed correct after @CamSoper's dispute (GitHub API re-check). One figure (Crossplane v2.3.4) remains unconfirmed, and three newly added pull-quotes could not be confirmed from the pages they cite.
coherence	MEDIUM	The new summary table's "drop-in HCL compatibility" contradicts the post's own OpenTofu section, which says "high, but not full, compatibility."
editorial balance	MEDIUM	Structurally sound — every table gives all ten tools a symmetric row — but the new AI-agents FAQ frames HCL tooling as a deficit rather than a tradeoff.

Investigation log

Cross-sibling reads: not run (not in a templated section)
External claim verification: 48 of 78 claims verified (8 unverifiable, 6 contradicted, 5 framing-drift) · 4 specialists (numerical, cross-reference, capability, framing); 0 cross-specialist corroborations · routed: 0 inline, 27 Pass 1, 4 Pass 2 (verified 0, contradicted 0, unverifiable 4), 47 Pass 3 (verified 28, contradicted 10, unverifiable 9).
Cited-claim spot-checks: 4 of 4 cited claims fetched and compared
Frontmatter sweep: ran on body + meta_desc
Temporal-trigger sweep: ran (recency words present in diff; spot-check in-review)
Code execution: not run (no static/programs/ change)
Code-examples checks: not run (no fenced code blocks in content files)
Editorial-balance pass: ran (12 H2 sections, 2 flags fired)

🚨 Outstanding	⚠️ Low-confidence	💡 Pre-existing	✅ Resolved
2	18	2	3

🔍 Verification trail

78 claims extracted · 48 verified · 8 unverifiable · 6 contradicted · 5 framing-drift

L4 in content/blog/infrastructure-as-code-tools/index.md "updated: 2026-07-31" → ➖ not-a-claim (evidence: This is frontmatter metadata (an "updated" date field) in the blog post's own file, not a falsifiable third-party assertion. It reflects the PR author's own publishing metadata, matching today's date (2026-07-31).; source: content/blog/infrastructure-as-code-tools/index.md)
L4-5 in content/blog/infrastructure-as-code-tools/index.md "This page was last updated/modified on 2026-07-31." → ➖ not-a-claim (evidence: The file's own frontmatter states updated: 2026-07-31 and lastmod: 2026-07-31, matching the claim exactly. This is simply a restatement of the page's own metadata field, not an independently falsifiable external assertion.; source: repo:content/blog/infrastructure-as-code-tools/index.md (frontmatter lines 4-5))
L5 in content/blog/infrastructure-as-code-tools/index.md "lastmod: 2026-07-31" → ➖ not-a-claim (evidence: This is a frontmatter metadata field (lastmod date) recording when the blog post's own content was last modified, not a falsifiable external assertion. It matches today's date context and is author-controlled metadata, not a checkable fact.; source: content/blog/infrastructure-as-code-tools/index.md (frontmatter, line 5))
L7 in content/blog/infrastructure-as-code-tools/index.md "This guide compares 10 IaC tools for 2026 on pricing, licensing, release cadence, and AI-agent readiness." → ➖ not-a-claim (evidence: This is a self-description of the blog post's own scope and structure (how many tools it covers, what criteria it uses) rather than a falsifiable third-party assertion; it's the author describing their own content design.; source: content/blog/infrastructure-as-code-tools/index.md (self-referential positioning statement))
L7 in content/blog/infrastructure-as-code-tools/index.md "Pulumi, Terraform, OpenTofu, AWS CDK, AWS CloudFormation, Azure ARM, Azure Bicep, Google Cloud Infrastructure Manager, Kubernetes YAML, and Crossplane are comp…" (also L37) → 🤷 unverifiable (framing: Cited source is unrelated to the claim's content (a package index vs. a 2026 tool-comparison analysis); page was fetched successfully but does not address the…; evidence: The cited URL is just the Pulumi Registry packages index listing (an alphabetical list of provider packages like AWS, Azure, etc.) and does not contain any content comparing Pulumi, Terraform, OpenTofu, AWS CDK, CloudFormation, ARM, Bicep,…; source: https://www.pulumi.com/registry/packages.md)
L37 in content/blog/infrastructure-as-code-tools/index.md "As of July 2026, Pulumi's package registry lists 308 unique packages, comprising first-party providers, bridged Terraform providers, and community components." → 🤷 unverifiable (evidence: No authoritative Pulumi source publishes an exact current count of registry packages. The registry page and community-packages list confirm many packages exist and continue to grow, but no page states "308" as a total; the closest historic…; source: WebSearch ran query "Pulumi package registry number of packages 2026"; top results (pulumi.com/registry/, pulumi.com/blog/introducing-pulumi-registry/) didn't state a current total count; intuition: A precise, non-round number like "308" for a live, constantly-changing registry count is hard to verify and looks like…)
L41-42 in content/blog/infrastructure-as-code-tools/index.md "In 2026, leading organizations are treating infrastructure as software, complete with testing, version control, code reviews, and continuous integration." → ➖ not-a-claim (evidence: This is a general positioning/opinion statement about industry trends ("leading organizations are treating infrastructure as software"), not a falsifiable, checkable assertion tied to a specific source, product feature, or statistic.; source: content/blog/infrastructure-as-code-tools/index.md (author's own framing/positioning statement))
L131 in content/blog/infrastructure-as-code-tools/index.md "Google Cloud Infrastructure Manager uses HCL (Terraform-based) and supports Google Cloud." (also L139) → ✅ verified (evidence: Google Cloud Infrastructure Manager (Config Manager) is Google's managed service that provisions and manages Google Cloud resources using standard Terraform configuration files (HCL), and it only manages Google Cloud resources — consistent…; source: Google Cloud documentation on Infrastructure Manager (cloud.google.com/infrastructure-manager/docs/overview) — general knowledge confirmed by table's internal consistency in content/blog/infrastructure-as-code-tools/index.md)
L131-142 in content/blog/infrastructure-as-code-tools/index.md "Among dedicated IaC tools, Terraform is the most-starred at roughly 49,300 stars, followed by OpenTofu at roughly 29,600, Pulumi at roughly 25,500, AWS CDK at…" (also L169-180, L201-202, L503-504, L557-558, L696-697, L787-788, L848-849, L882-883, L1479-1494) → ✅ verified (evidence: Independent sources confirm OpenTofu's star count matches the claim precisely: GitHub itself shows "Star 29.6k" for opentofu/opentofu, and a June 2026 third-party source notes "With over 29,000 GitHub stars as of June 2026." This corrobora…; source: https://github.com/opentofu/opentofu (Star 29.6k); https://scalr.com/learning-center/what-is-opentofu (June 2026, "over 29,000 GitHub stars"))
L131 in content/blog/infrastructure-as-code-tools/index.md "Terraform's language/approach is HCL, HashiCorp's DSL." (also L134) → ✅ verified (evidence: Official HashiCorp docs confirm Terraform's configuration language is based on HCL, HashiCorp's own domain-specific language: "Terraform's configuration language is based on a more general language called HCL."; source: https://developer.hashicorp.com/terraform/language/syntax/configuration)
L131 in content/blog/infrastructure-as-code-tools/index.md "Crossplane uses YAML / Kubernetes CRDs and is multi-cloud, orchestrated through Kubernetes." (also L140-141) → ✅ verified (evidence: Multiple independent sources confirm Crossplane's architecture: "It uses Kubernetes Custom Resource Definitions (CRDs) to ensure that users interact with infrastructure uniformly, regardless of the underlying provider" and it acts as "a un…; source: https://komodor.com/learn/what-is-crossplane-pros-cons-and-a-quick-tutorial/; https://www.cloudthat.com/resources/blog/multi-cloud-application-deployment-and-management-using-crossplane)
L131 in content/blog/infrastructure-as-code-tools/index.md "| Tool | Language / approach | Clouds supported | License | Latest stable release (Jul 2026) | Best for |" → ✅ verified (evidence: gh release list confirms v3.255.0 (2026-07-28T13:37:19Z) is tagged "Latest" for pulumi/pulumi, matching the table's claim of v3.255.0 (2026-07-28) as latest stable release as of Jul 2026.; source: gh release list -R pulumi/pulumi --limit 5)
L131 in content/blog/infrastructure-as-code-tools/index.md "Terraform's latest stable release as of the table's reference date is v1.15.8, released 2026-07-08." (also L133-134, L172, L201-202, L503-504) → ✅ verified (evidence: GitHub Releases API for hashicorp/terraform confirms tag_name "v1.15.8", prerelease:false, published_at "2026-07-08T17:45:09Z" — an exact match to the claim's stated version and date, and the table's own "as of Jul 2026" reference date fra…; source: gh api repos/hashicorp/terraform/releases/tags/v1.15.8)
L131 in content/blog/infrastructure-as-code-tools/index.md "Terraform is licensed under BUSL-1.1." (also L133-134, L172, L201-202, L503-504) → ✅ verified (evidence: HashiCorp officially relicensed Terraform from MPL 2.0 to Business Source License (BUSL) 1.1 in August 2023, a widely reported and documented change; the doc's statement "Terraform | BUSL-1.1 (not OSI-approved) | HashiCorp (an IBM company)…; source: Public knowledge: HashiCorp Terraform license change announcement (Aug 10, 2023), corroborated by repo content itself referencing "its 2023 licensing change to BUSL-1.1 (no longer open source)".)
L135 in content/blog/infrastructure-as-code-tools/index.md "| AWS CDK | TypeScript, Python, Java, C#, Go (compiles to CloudFormation) | AWS only | Apache-2.0 | v2.262.2 (2026-07-29) |…" → ✅ verified (evidence: GitHub releases page for aws/aws-cdk shows "aws-cdk-automation released this · 29 Jul 14:24 · v2.262.2" as the latest release, matching the claimed version and date exactly.; source: https://github.com/aws/aws-cdk/releases)
L138 in content/blog/infrastructure-as-code-tools/index.md "| Azure Bicep | Bicep DSL (compiles to ARM JSON) | Azure only | MIT | v0.46.1 (2026-07-30) | Azure deployments requiring improved readability…" → ✅ verified (evidence: Disputed by @CamSoper on 2026-08-04. Re-checked via gh api repos/Azure/bicep/releases/latest: tag v0.46.1, published 2026-07-30T23:35:23Z, prerelease: false — the table's version and date are correct. The original check's source page evidently returned a stale snapshot (v0.45.6 / v0.45.15) predating this release.; source: gh api repos/Azure/bicep/releases/latest)
L142 in content/blog/infrastructure-as-code-tools/index.md "| OpenTofu | HCL (community-governed Terraform fork) | AWS, Azure, Google Cloud, and the Terraform provider ecosystem | MPL-2.0 | v1.12.5 (2026…" → ✅ verified (evidence: GitHub shows the OpenTofu v1.12.5 release was published "21 Jul 11:39", matching the claim's "v1.12.5 (2026-07-21)"; MPL-2.0 license and community governance are also independently confirmed by other sources.; source: https://github.com/opentofu/opentofu/releases/tag/v1.12.5)
L144 in content/blog/infrastructure-as-code-tools/index.md "The version and release-date figures in the comparison table were checked against each project's GitHub releases on 2026-07-31." → ➖ not-a-claim (evidence: This is a statement describing the author's own fact-checking methodology/process for the blog post table (an editorial attribution note), not a falsifiable third-party assertion that can be independently verified via an external source.; source: content/blog/infrastructure-as-code-tools/index.md L144)
L148 in content/blog/infrastructure-as-code-tools/index.md "Tool cost is rarely just the license. Here's what each option actually costs to run, based on published pricing as of July 2026:" → 🤷 unverifiable (evidence: The claim asserts pricing data "as of July 2026," a future-dated snapshot claim that cannot be checked against any current source; pricing pages change over time and no archived snapshot for that date exists to confirm.; source: content/blog/infrastructure-as-code-tools/index.md (self-dated future claim, no independent source available))
L152 in content/blog/infrastructure-as-code-tools/index.md "Pulumi's Individual plan is free forever, supports 1 user, unlimited stacks, 500 workflow minutes per month, and 5 million Pulumi Neo tokens per month." (also L1483) → ✅ verified (evidence: content/pricing/_index.md lists the Individual tier as "$0"/"forever" with features: "1 user", "Unlimited projects, stacks, and environments", "Up to 500 workflow minutes", and "Pulumi Neo with 5M free tokens/month" — matching all figures…; source: repo:content/pricing/_index.md)
L153 in content/blog/infrastructure-as-code-tools/index.md "HCP Terraform's Standard tier is priced at $0.47 per resource per month, and the Premium tier is priced at $0.99 per resource per month." (also L1483) → ✅ verified (evidence: HashiCorp's official pricing page shows Premium "Starting at $0.99 per month per resource," and multiple independent sources confirm Standard is $0.47 and Premium is $0.99 per resource per month.; source: https://www.hashicorp.com/en/pricing)
L154-160 in content/blog/infrastructure-as-code-tools/index.md "AWS CDK, AWS CloudFormation, Azure ARM, Azure Bicep, Google Cloud Infrastructure Manager, Kubernetes YAML, Crossplane, and OpenTofu carry no separate tool fee." → ✅ verified (evidence: The blog's own pricing table (which this claim summarizes) lists each of these tools as free/open-source or "included" with their respective cloud platform, e.g. "AWS CDK | No tool fee — free and open source", "Azure ARM | No tool fee — in…; source: content/blog/infrastructure-as-code-tools/index.md lines 154-161 (Pricing at a glance table))
L163 in content/blog/infrastructure-as-code-tools/index.md "Spacelift and env0 price by usage-based or per-resource quotes rather than published tiers." → ❌ contradicted (framing: Claim asserts pricing is entirely usage-based/per-resource quotes "rather than published tiers," but both vendors publish named entry-level tiers with fixed st…; evidence: (escalated from pass1) Both vendors publish named tiers with starting prices, not just usage-based quotes: Spacelift lists a Free plan and "Starter plan begins at $399/month for 10 users," with only higher tiers (Business/Enterprise) using…; source: https://zoftwarehub.com/products/spacelift/pricing ; https://www.getapp.com/it-management-software/a/env0/pricing/ ; https://www.saasworthy.com/product/env0/pricing)
L169-180 in content/blog/infrastructure-as-code-tools/index.md "Terraform is governed by HashiCorp, which is described as an IBM company." → ✅ verified (evidence: IBM completed its acquisition of HashiCorp in February 2025; per LegalClarity, "Its website identifies it as 'HashiCorp, an IBM Company,' and the full product lineup remains intact," confirming HashiCorp (governor of Terraform) is now desc…; source: https://legalclarity.org/who-owns-hashicorp-ibms-acquisition-explained/)
L169 in content/blog/infrastructure-as-code-tools/index.md "| Tool | License | Governing body | Latest version | Released | GitHub stars |" → ➖ not-a-claim (evidence: This is a table header row (column labels) from a blog comparison table, not a factual assertion that can be verified independently.; source: content/blog/infrastructure-as-code-tools/index.md L169)
L171 in content/blog/infrastructure-as-code-tools/index.md "Pulumi's governing body is Pulumi Corporation." → ✅ verified (evidence: Wikipedia confirms: "Pulumi Corporation is a software company based in Seattle, Washington" that "develops an open-source infrastructure-as-code software," consistent with Pulumi Corporation being the company/governing body behind the Pulu…; source: https://en.wikipedia.org/wiki/Pulumi)
L171 in content/blog/infrastructure-as-code-tools/index.md "| Pulumi | Apache-2.0 | Pulumi Corporation | v3.255.0 | 2026-07-28 | ~25.5k |" → ✅ verified (evidence: GitHub confirms pulumi/pulumi released v3.255.0 on 2026-07-28 ("github-actions released this · 28 Jul 13:37 · v3.255.0") and the repo shows "Star 25.5k", matching the claim's version, date, and ~25.5k star count. Pulumi is Apache-2.0 licen…; source: https://github.com/pulumi/pulumi/releases)
L171-172 in content/blog/infrastructure-as-code-tools/index.md "Terraform has approximately 49,300 GitHub stars." → ✅ verified (evidence: Multiple current (July 2026) sources put hashicorp/terraform's GitHub star count around 49k: a release tracker shows "Stars49.2K" and GitHub's own Issues page shows "Star 49.3k", closely matching the claimed ~49,300.; source: https://github.com/hashicorp/terraform/issues (Star 49.3k); https://releasealert.dev/github/hashicorp/terraform (Stars 49.2K))
L173 in content/blog/infrastructure-as-code-tools/index.md "| AWS CDK | Apache-2.0 | AWS | v2.262.2 | 2026-07-29 | ~12.9k |" → ✅ verified (framing: Version/date/license/maintainer match exactly; star count (~12.9k claimed vs ~12.6k observed) is an approximate figure consistent with normal star-count growth…; evidence: GitHub releases page confirms: "aws-cdk-automation released this · 29 Jul 14:24 · v2.262.2" matching the 2026-07-29 date and version exactly. License is Apache-2.0 and maintainer is AWS per the repo. Star count on aws/aws-cdk shows "Star 1…; source: https://github.com/aws/aws-cdk/releases (v2.262.2, 29 Jul); https://github.com/aws/aws-cdk/graphs/contributors (Star 12.6k))
L176 in content/blog/infrastructure-as-code-tools/index.md "| Azure Bicep | MIT | Microsoft | v0.46.1 | 2026-07-30 | ~3.6k |" → ✅ verified (evidence: Disputed by @CamSoper on 2026-08-04; same re-check as L138 — Azure/bicep's latest release is v0.46.1, published 2026-07-30T23:35:23Z, prerelease: false. The license, governing body, and ~3.6k star count were already correct.; source: gh api repos/Azure/bicep/releases/latest)
L178-179 in content/blog/infrastructure-as-code-tools/index.md "| Kubernetes | Apache-2.0 | CNCF | v1.36.3 | 2026-07-23 | ~124.1k |" → ✅ verified (evidence: Kubernetes 1.36.3 was released around July 22-23, 2026 ("Kubernetes 1.36.3 is the latest supported version (July 2026)" per eosl.date; versionlog.com lists "1.36.3 22 Jul 2026"), matching the claimed version and date. Star count (~124.1k)…; source: https://eosl.date/eol/product/kubernetes/ ; https://versionlog.com/kubernetes/ ; https://github.com/topics/kubernetes)
L180 in content/blog/infrastructure-as-code-tools/index.md "| OpenTofu | MPL-2.0 | Linux Foundation | v1.12.5 | 2026-07-21 | ~29.6k |" → ✅ verified (evidence: GitHub confirms OpenTofu is MPL-2.0 licensed, hosted by Linux Foundation, with releases page showing "v1.12.5 Latest" and repo star count listed as "Star 29.6k" — matching the claim's version (v1.12.5) and star count (~29.6k). The release…; source: https://github.com/opentofu/opentofu/releases and https://github.com/opentofu/opentofu/pkgs/container/opentofu)
L182 in content/blog/infrastructure-as-code-tools/index.md "Terraform's move to the Business Source License occurred in 2023." → ✅ verified (evidence: HashiCorp publicly announced on August 10, 2023 that Terraform (and other HashiCorp products) would move from MPL-2.0 to the Business Source License (BUSL 1.1), which prompted the OpenTofu fork from Terraform's last MPL-2.0 release — match…; source: Well-documented public event: HashiCorp blog announcement, Aug 10, 2023, "Hashicorp adopts Business Source License"; corroborated internally at content/blog/infrastructure-as-code-tools/index.md lines 182, 507.)
L191 in content/blog/infrastructure-as-code-tools/index.md "| You have deep existing Terraform/HCL expertise and workflows | Terraform | Largest ecosystem and community knowledge base, despite the BUSL-1.1 licensing tra…" → ✅ verified (framing: Source confirms both the BUSL-1.1 licensing change and Terraform's large ecosystem/registry independently; the claim's comparative framing ("largest... despite…; evidence: HashiCorp confirmed changing Terraform's license to BUSL 1.1 in 2023, and Terraform is widely described as having a large registry/community ecosystem: "Terraform has a large ecosystem of publicly available providers" and its registry serv…; source: WebSearch ran queries "Terraform BUSL-1.1 license change" and "Terraform largest ecosystem community modules registry")
L192 in content/blog/infrastructure-as-code-tools/index.md "OpenTofu offers drop-in HCL compatibility." → 🌀 framing-drift (framing: overclaim — article body says "high, but not full, compatibility... with some divergence emerging"; summary table claims unqualified "drop-in HCL compatibility"; evidence: The same blog post's own detailed OpenTofu section (line 886) qualifies this claim, stating OpenTofu provides "high, but not full, compatibility with Terraform" and notes "some divergence emerging as the projects evolve independently" (lin…; source: content/blog/infrastructure-as-code-tools/index.md (self-referential, lines 886, 891 vs. 192); intuition: Absolute "drop-in" framing conflicts with the article's own more careful/hedged language elsewhere in the same post.)
L197 in content/blog/infrastructure-as-code-tools/index.md "| You're deploying AI agents to manage infrastructure changes | Pulumi | Real languages give agents testable, reviewable code to reason about, rather than opaq…" → ✅ verified (framing: Source describes Neo's code-generation and lifecycle-automation capabilities in general languages; claim's narrower framing (agents reasoning about testable/re…; evidence: The /product/neo/ page confirms Neo is an AI infrastructure agent that "can create Pulumi programs in TypeScript, Python, Go, and more from a simple prompt, following Pulumi best practices" and handles "the full infrastructure lifecycle" w…; source: content/product/neo.md)
L212 in content/blog/infrastructure-as-code-tools/index.md "Pulumi provides native cloud provider SDKs enabling same-day feature access across 170+ providers." (also L483) → 🌀 framing-drift (framing: overclaim — source scopes "same-day support for new features" to native providers (AWS/Azure/GCP/K8s), claim applies it across "170+ providers" in the registry; evidence: Same-day feature access is a documented characteristic specifically of Pulumi's Native Providers (AWS Native, Azure Native, Google Cloud Native), generated automatically from cloud provider APIs: "Native providers are built directly and au…; source: https://www.pulumi.com/blog/pulumiup-native-providers/ and https://www.pulumi.com/blog/any-terraform-provider/)
L487-490 in content/blog/infrastructure-as-code-tools/index.md "Yarin Miran, Senior Software Engineer at Wiz, said: 'We use Pulumi widely at Wiz. It enabled our product to support multi-cloud and to scale quickly - scaling…" → 🤷 unverifiable (evidence: The pre-fetched body of https://www.pulumi.com/case-studies/wiz/ is truncated to generic site navigation/header content and does not include the quoted text attributed to Yarin Miran. Cannot confirm the quote appears on the cited page from…; source: https://www.pulumi.com/case-studies/wiz/)
L497 in content/blog/infrastructure-as-code-tools/index.md "Unity, Snowflake, and Starburst reported significant productivity improvements, specifically 80-90% deployment time reductions, when adopting programming langu…" → ❌ contradicted (framing: Claim invents a uniform "80-90%" range and attributes it to all three companies; actual reported figures are a 112x/99% reduction for Starburst, an unquantifie…; evidence: Pulumi's own case studies report far more dramatic and differently-framed figures than "80-90% deployment time reductions": Starburst reports a "112x faster deployment time" (a 99% reduction, two weeks to three hours), and Snowflake report…; source: https://www.pulumi.com/case-studies/starburst/ ; https://www.pulumi.com/case-studies/snowflake/ ; https://www.pulumi.com/case-studies/unity/; intuition: Suspiciously round shared range (80-90%) attributed uniformly to three different companies whose actual reported number…)
L504 in content/blog/infrastructure-as-code-tools/index.md "Latest stable release: v1.15.8 (2026-07-08)" → ✅ verified (evidence: HashiCorp's official GitHub release page states "1.15.8 (July 8, 2026)" as the release, matching the claimed version and date, and it is confirmed as the latest stable release (next release 1.16.0 was only alpha/beta at the time).; source: https://github.com/hashicorp/terraform/releases/tag/v1.15.8)
L507 in content/blog/infrastructure-as-code-tools/index.md "Terraform's licensing change to BUSL-1.1 occurred in 2023." → ✅ verified (evidence: HashiCorp publicly announced on August 10, 2023 that Terraform (and other HashiCorp products) would move from MPL 2.0 to the Business Source License (BUSL-1.1) starting with future releases; this is the same event the blog post references…; source: content/blog/infrastructure-as-code-tools/index.md (lines 182, 507); public knowledge of HashiCorp's August 10, 2023 BUSL license change announcement)
L558 in content/blog/infrastructure-as-code-tools/index.md "Latest stable release: v2.262.2 (2026-07-29)" → ❌ contradicted (evidence: No tool in this space has version v2.262.2. Terraform's latest stable is reported as "Terraform 1.15.x (1.15.8 as of July 2026)" and Pulumi's latest releases are v3.254.0/v3.255.0 (per GitHub releases page, July 2026). The claimed version…; source: WebSearch ran query "Terraform latest stable release v2.262.2" and "Pulumi CLI latest release version July 2026"; top results show Terraform at 1.15.x and Pulumi at v3.254.0/v3.255.0, not v2.262.2; intuition: Version string v2.262.2 doesn't match any known tool's actual release numbering scheme found in search (Terraform is 1.…)
L565 in content/blog/infrastructure-as-code-tools/index.md "AWS CDK supports TypeScript, Python, Java, C#, and JavaScript with full IDE integration, and Go is available in Developer Preview." → ❌ contradicted (framing: shifted — claim states Go is currently "Developer Preview" (true circa 2021-2022) but current AWS docs say Go "is considered stable" and fully supported, so th…; evidence: (escalated from pass1) Current AWS CDK documentation states Go has graduated from Developer Preview: "Go is a fully-supported client language for the AWS Cloud Development Kit (AWS CDK) and is considered stable." The v2 languages page also…; source: https://docs.aws.amazon.com/cdk/v2/guide/work-with-cdk-go.html; intuition: Go's Developer Preview status for AWS CDK ended years ago; this claim appears to use stale information.)
L697 in content/blog/infrastructure-as-code-tools/index.md "Latest stable release: v0.46.1 (2026-07-30)" → ✅ verified (evidence: Disputed by @CamSoper on 2026-08-04; same re-check as L138/L176 — this is the Azure Bicep section header, and Azure/bicep's latest release is genuinely v0.46.1 (2026-07-30T23:35:23Z, non-prerelease).; source: gh api repos/Azure/bicep/releases/latest)
L741 in content/blog/infrastructure-as-code-tools/index.md "Google Cloud Infrastructure Manager uses Terraform configurations to automate deployment and management of Google Cloud infrastructure resources." → ✅ verified (evidence: GoogleCloudPlatform's own sample repo README states: "This directory contains a sample implementation of a Cloud Deploy Custom Target for deploying Google Cloud infrastructure resources via Terraform using [Infrastructure Manager](https://…; source: gh api repos/GoogleCloudPlatform/cloud-deploy-samples/contents/custom-targets/infrastructure-manager/README.md)
L788 in content/blog/infrastructure-as-code-tools/index.md "Latest stable release: Kubernetes v1.36.3 (2026-07-23)" → ✅ verified (evidence: GitHub releases page for kubernetes/kubernetes shows v1.36.3 as the latest release, "k8s-release-robot released this · 23 Jul 00:35 · v1.36.3", consistent with the claimed 2026-07-23 date; a third-party EOL tracker also states "Kubernetes…; source: https://github.com/kubernetes/kubernetes/releases)
L822 in content/blog/infrastructure-as-code-tools/index.md "The Kubernetes YAML code examples in this article use the nginx:1.27 container image." (also L1122) → ✅ verified (evidence: Both cited lines in the article's Kubernetes YAML examples show image: nginx:1.27 (line 822 in the Kubernetes-native deployment example and line 1122 in the GitOps-style deployment example).; source: repo:content/blog/infrastructure-as-code-tools/index.md (lines 822, 1122))
L849 in content/blog/infrastructure-as-code-tools/index.md "Latest stable release: v2.3.4 (2026-07-23)" → 🤷 unverifiable (evidence: No search result identifies which tool "v2.3.4 (2026-07-23)" refers to, nor confirms such a release exists; Pulumi's own CLI version tracked by Wikipedia is 3.243.0 (May 22, 2026), not v2.3.4, so this figure does not correspond to Pulumi i…; source: WebSearch ran query "v2.3.4 latest stable release 2026-07-23"; top results didn't address the claim; intuition: Suspiciously precise version/date pair with no identifiable tool name attached; smells like a fabricated or template-fi…)
L886 in content/blog/infrastructure-as-code-tools/index.md "OpenTofu is maintained by the Linux Foundation." → ✅ verified (evidence: Linux Foundation's own press release confirms: "the Linux Foundation announced the formation of OpenTofu, an open source alternative to Terraform's widely used infrastructure as code provisioning tool," and OpenTofu's own site states it is…; source: https://www.linuxfoundation.org/press/announcing-opentofu; https://opentofu.org/)
L886 in content/blog/infrastructure-as-code-tools/index.md "OpenTofu provides high but not full compatibility with Terraform." → ✅ verified (evidence: (escalated from pass1) Multiple sources confirm OpenTofu is highly but not fully compatible with Terraform: it offers "100% backwards compatibility with Terraform versions prior to 1.6" and reads Terraform state/providers directly, but sou…; source: https://spacelift.io/blog/opentofu-vs-terraform)
L886 in content/blog/infrastructure-as-code-tools/index.md "OpenTofu emerged as a fork of Terraform v1.5.x following HashiCorp's license change." → ✅ verified (evidence: OpenTofu's own RFC states: "This is the result of forking Terraform ~1.5.5" — confirming OpenTofu forked from the Terraform v1.5.x line after HashiCorp's August 2023 BSL license change, matching the claim exactly.; source: gh search code --owner opentofu "1.5" (opentofu/opentofu:rfc/20240529-OpenTofu-Specific-Code-Override.md))
L924 in content/blog/infrastructure-as-code-tools/index.md "Puppet is licensed under Apache 2.0." (also L969, L1008, L1051) → ✅ verified (evidence: The puppetlabs/puppet GitHub repository's LICENSE file (decoded) begins "Apache License, Version 2.0, January 2004," confirming Puppet's core is licensed under Apache 2.0.; source: gh api repos/puppetlabs/puppet/contents/LICENSE)
L1122 in content/blog/infrastructure-as-code-tools/index.md "image: nginx:1.27" → ➖ not-a-claim (evidence: This is a version pin in an illustrative code example (image: nginx:1.27) within a blog post about IaC tools, not a factual assertion about Pulumi or a third party that requires verification.; source: content/blog/infrastructure-as-code-tools/index.md L1122)
L1152 in content/blog/infrastructure-as-code-tools/index.md "Checkov supports Terraform, CloudFormation, Kubernetes, Helm, and ARM templates." → ✅ verified (framing: Source lists a broader set of supported frameworks; the claim names a subset of them, which is entailed.; evidence: Checkov's official README states: "Scans Terraform, Terraform Plan, Terraform JSON, CloudFormation, AWS SAM, Kubernetes, Helm, Kustomize, Dockerfile, Serverless framework, Ansible, Bicep, ARM, and OpenTofu template files." This confirms Te…; source: gh search code --repo bridgecrewio/checkov "Helm" "ARM" (README.md))
L1152 in content/blog/infrastructure-as-code-tools/index.md "Checkov provides over 1000 built-in policies covering CIS benchmarks, PCI DSS, and GDPR compliance." → ✅ verified (framing: Source lists a broader set of frameworks (CIS, PCI-DSS, HIPAA, SOC2, NIST, GDPR across different sources); claim picks a subset (CIS, PCI DSS, GDPR) which is a…; evidence: Multiple independent sources confirm Checkov's "over 1,000 built-in policies" figure covering CIS benchmarks and PCI DSS; e.g. "with over 1,000 built-in policies covering major cloud providers... and compliance standards like CIS Benchmark…; source: https://devsecopsschool.com/blog/comprehensive-tutorial-on-checkov-in-devsecops/ ; https://spacelift.io/blog/what-is-checkov)
L1156 in content/blog/infrastructure-as-code-tools/index.md "TFLint is licensed under MPL 2.0." → ✅ verified (framing: Source shows dual licensing (MPL 2.0 core + BUSL 1.1 for vendored terraform internals); claim's assertion that TFLint is MPL 2.0 licensed is the core/primary l…; evidence: TFLint's repo README badge states "License: MPL 2.0 + BUSL 1.1" and its main LICENSE is MPL 2.0 (CHANGELOG notes "Change license: MIT -> MPL 2.0"); the BUSL portion applies only to vendored Terraform core code, not the TFLint tool itself.; source: gh search code --repo terraform-linters/tflint LICENSE (README.md badge: "License: MPL 2.0 + BUSL 1.1"))
L1285-1288 in content/blog/infrastructure-as-code-tools/index.md "Joe Duffy's quote about Pulumi Neo is sourced from a PR Newswire press release titled introducing Pulumi Neo as the industry's first AI-powered platform engine…" → ✅ verified (evidence: (escalated from pass1) The PR Newswire release "Introducing Pulumi Neo, the Industry's First AI-Powered Platform Engineer" (Sept 16, 2025) quotes: "'In under a year, AI has completely reshaped how we build applications. With Pulumi Neo, pl…; source: https://www.prnewswire.com/news-releases/introducing-pulumi-neo-the-industrys-first-ai-powered-platform-engineer-302556718.html)
L1285-1288 in content/blog/infrastructure-as-code-tools/index.md "Joe Duffy, Founder & CEO of Pulumi, said: 'In under a year, AI has completely reshaped how we build applications. With Pulumi Neo, platform engineering is now…" → 🤷 unverifiable (evidence: The pre-fetched body is only the generic PR Newswire navigation/boilerplate content; the truncated body does not contain the actual press release text or Joe Duffy's quote, so the exact quote cannot be confirmed from this fetch.; source: https://www.prnewswire.com/news-releases/introducing-pulumi-neo-the-industrys-first-ai-powered-platform-engineer-302556718.html)
L1311 in content/blog/infrastructure-as-code-tools/index.md "Atlassian Bitbucket's rapid 2-day migration was enabled by existing Python expertise on the team, well-structured Terraform code, and straightforward infrastru…" → 🌀 framing-drift (framing: source credits the automatic conversion tool for the fast timeline; claim substitutes unsupported reasons ("well-structured Terraform code, straightforward inf…; evidence: The case study confirms the 2-day timeline and Python-familiarity rationale: "Because all the developers on Bitbucket write Python, they can use all the skills they already have" and "I spent one day using the automatic conversion tool...…; source: https://www.pulumi.com/case-studies/atlassian/)
L1311 in content/blog/infrastructure-as-code-tools/index.md "Atlassian Bitbucket completed a full migration from Terraform to Pulumi in 2 days using automated conversion tools." → 🌀 framing-drift (framing: overclaim — source describes converting a single pipeline/project in two days; claim broadens this to "a full migration" of Bitbucket's Terraform usage to Pulu…; evidence: The Pulumi case study quotes Mike: "I spent one day using the automatic conversion tool and then just spent some time making sure everything looked right... In two days, I was all set." This describes converting one specific CI/CD pipeline…; source: https://www.pulumi.com/case-studies/atlassian/)
L1315-1318 in content/blog/infrastructure-as-code-tools/index.md "Matt Stephenson, Senior Principal Software Engineer at Starburst, said: 'When we did it with Terraform, it took two weeks to do [infrastructure deployments]. N…" → 🤷 unverifiable (evidence: The pre-fetched body of the cited case-study URL is truncated to generic site navigation/header content and does not contain the quoted passage attributed to Matt Stephenson, so the specific quote could not be confirmed against the source.; source: https://www.pulumi.com/case-studies/starburst/)
L1460 in content/blog/infrastructure-as-code-tools/index.md "Atlassian Bitbucket completed its migration from Terraform to Pulumi's Python-based IaC in just 2 days using migration tools." → ✅ verified (evidence: Pulumi's Atlassian case study quotes Bitbucket engineer Mike: "I spent one day using the automatic conversion tool and then just spent some time making sure everything looked right... In two days, I was all set," referring to migrating fro…; source: https://www.pulumi.com/case-studies/atlassian/)
L1461 in content/blog/infrastructure-as-code-tools/index.md "Starburst achieved 112x faster deployments after incrementally migrating from Terraform to Pulumi." → ✅ verified (framing: Claim's wording "incrementally migrating from Terraform to Pulumi" matches the case study's description of switching from Terraform to Pulumi; "112x faster dep…; evidence: Pulumi's official Starburst case study states: "Starburst transformed its infrastructure provisioning and management processes by switching from Terraform to Pulumi, resulting in a 112x faster deployment time and significant cost savings,"…; source: https://www.pulumi.com/case-studies/starburst/)
L1462 in content/blog/infrastructure-as-code-tools/index.md "Multiple organizations report 80-90% deployment time reductions after moving to programming language-based IaC approaches." (also L1505) → 🌀 framing-drift (framing: Claim compresses varied figures (80% Unity, 99% Starburst, 70% SANS) into a uniform "80-90%" range attributed to "multiple organizations," but at least one cit…; evidence: The blog itself states "Organizations like Unity, Snowflake, and Starburst have reported significant productivity improvements (80-90% deployment time reductions)," but Pulumi's own case studies show Unity reported "80% reduction in deploy…; source: https://www.pulumi.com/case-studies/starburst/ and https://www.pulumi.com/case-studies/unity/; intuition: The specific 80-90% range doesn't match the actual spread of reported figures (70%-99%) across Pulumi's own case studie…)
L1479-1481 in content/blog/infrastructure-as-code-tools/index.md "AWS CDK, CloudFormation, Azure ARM, Azure Bicep, and Google Cloud Infrastructure Manager carry no separate tool fee, and you pay only for the cloud resources y…" → ✅ verified (evidence: The article's own pricing table lists AWS CDK, AWS CloudFormation, Azure ARM, Azure Bicep, and Google Cloud Infrastructure Manager each as "No tool fee — free/included..., pay only for provisioned ... resources," consistent with well-docum…; source: content/blog/infrastructure-as-code-tools/index.md lines 154-158 (internal pricing table); general vendor pricing knowledge for AWS CloudFormation/CDK, Azure ARM/Bicep, and Google Cloud Infrastructure Manager)
L1479-1494 in content/blog/infrastructure-as-code-tools/index.md "Open-source tools including OpenTofu, Kubernetes, and Crossplane are free to run, though users still cover their own compute and cluster costs." → ✅ verified (evidence: OpenTofu, Kubernetes, and Crossplane are all genuinely open-source (MPL 2.0 / Apache 2.0 licensed) projects with no license fee; the claim's own text correctly notes users still pay for their own compute/cluster costs, which is standard fo…; source: General knowledge of OpenTofu/Kubernetes/Crossplane open-source licensing (Apache 2.0 / MPL 2.0), consistent with content/blog/infrastructure-as-code-tools/index.md L1481.)
L1483 in content/blog/infrastructure-as-code-tools/index.md "Pulumi Cloud's Individual plan is free forever for one user with unlimited stacks and projects, 500 workflow minutes per month, and 5 million Pulumi Neo tokens…" → ✅ verified (evidence: content/pricing/_index.md lists the Individual tier as "price: $0 / price_label: forever" with features "1 user", "Unlimited projects, stacks, and environments", "Up to 500 workflow minutes", and "Pulumi Neo with 5M free tokens/month" — ma…; source: repo:content/pricing/_index.md)
L1483 in content/blog/infrastructure-as-code-tools/index.md "HCP Terraform is free for up to 500 managed resources with one concurrent run, then moves to per-resource pricing of roughly $0.10 to $0.99 per resource per mo…" → ✅ verified (evidence: Multiple 2026 sources confirm: "Free – up to 500 managed resources per organization, 1 concurrent run" and paid tiers "Essentials – starts around $0.10 per managed resource per month... Standard – starts around $0.47... Premium – starts ar…; source: https://spacelift.io/blog/terraform-cloud-pricing)
L1483 in content/blog/infrastructure-as-code-tools/index.md "The pricing figures for Pulumi Cloud and HCP Terraform reflect published pricing as of July 2026." → ➖ not-a-claim (evidence: This is a self-disclosed temporal scoping statement within the blog post itself ("Figures reflect published pricing as of July 2026 and change periodically, so confirm current rates before budgeting"), not a third-party factual assertion.…; source: repo:content/pricing/_index.md)
L1487 in content/blog/infrastructure-as-code-tools/index.md "Pulumi Neo can propose changes, run previews, respond to failures, and open pull requests using Python, TypeScript, Go, C#, or Java." → ✅ verified (framing: Source confirms preview/failure-response/PR-creation capabilities generally (via native VCS integrations); claim narrows this to the specific supported SDK lan…; evidence: Pulumi Neo is confirmed via pulumi/pulumi changelogs to run previews (pulumi neo --debug-preview), respond to failures (--debug-update/--debug-preview flags "to investigate a failed update or preview"), and via pulumi/docs to open pu…; source: gh search code --owner pulumi "pulumi neo"; repo:pulumi/docs content/docs/integrations/version-control/custom-vcs.md)
L1487 in content/blog/infrastructure-as-code-tools/index.md "Tools built on general-purpose programming languages give AI agents the biggest advantage, because agents can read, generate, test, and refactor real code the…" → ✅ verified (evidence: The /product/neo/ page describes Neo as "the industry's first AI agent built from the ground up to execute, govern, and optimize complex cloud automation," capable of generating infrastructure code, debugging, and executing changes — consi…; source: repo:content/product/neo.md)
L1489 in content/blog/infrastructure-as-code-tools/index.md "Agents working in HCL with Terraform and OpenTofu lack access to native unit testing and IDE tooling." → ❌ contradicted (evidence: Both Terraform and OpenTofu have native unit testing support: "Both Terraform (v1.6+) and OpenTofu support native unit testing using HCL-based .tftest.hcl files, no additional language or framework required." IDE tooling is also widely ava…; source: https://spacelift.io/blog/terraform-test)
L1489 in content/blog/infrastructure-as-code-tools/index.md "'AI is accelerating software delivery' is becoming an operating assumption for engineering teams." → ➖ not-a-claim (evidence: This is a general industry-positioning statement/opinion about AI accelerating software delivery, not a specific, falsifiable factual assertion tied to a checkable source.; source: n/a)
L1491 in content/blog/infrastructure-as-code-tools/index.md "### What is the most popular infrastructure as code tool in 2026?" → ➖ not-a-claim (evidence: This is a section heading phrased as a question ("What is the most popular infrastructure as code tool in 2026?"), not a falsifiable factual assertion itself; it's a FAQ prompt structure, so there's nothing to verify at the heading level.; source: content/blog/infrastructure-as-code-tools/index.md, line 1491)
L1491-1493 in content/blog/infrastructure-as-code-tools/index.md "The GitHub star counts in the article were measured as of 2026-07-31." → 🤷 unverifiable (evidence: The claim states GitHub star counts were measured as of a future date (2026-07-31), which is a self-reported snapshot timestamp by the article's own author and cannot be independently verified against any external source — star counts cons…; source: not-applicable: future-dated self-reported measurement)
L1493 in content/blog/infrastructure-as-code-tools/index.md "By GitHub stars, Kubernetes leads the broader infrastructure ecosystem at roughly 124,100 stars." → ✅ verified (framing: Source shows ~123k (rounded), claim states a more precise 124,100 — consistent within normal star-count growth/rounding.; evidence: GitHub topics listing shows kubernetes/kubernetes at "Star 123k" as of a June 2026 update, which rounds to roughly the claimed 124,100 stars.; source: https://github.com/topics/kubernetes)
L1493 in content/blog/infrastructure-as-code-tools/index.md "Terraform's larger community reflects a decade of first-mover adoption." → ✅ verified (evidence: Terraform launched in July 2014 while Pulumi launched in 2018, giving Terraform roughly a decade of head start; multiple sources confirm "Terraform was launched by HashiCorp in 2014" and describe its subsequent community growth "to hundred…; source: https://en.wikipedia.org/wiki/Terraform_(software); https://build5nines.com/software-innovation-hashicorp-terraform-revolutionized-infrastructure-as-code-iac/)

Refresh this review — comment @claude #update-review. Say what you fixed, or which finding you dispute and why; both work in the same mention.
Ask for anything else — comment @claude with no hashtag (questions, one-off fixes). Leaves this review untouched.

Important

Please don't hide, resolve, or delete this comment! It breaks things!

📖 How pre-merge review works — the full lifecycle, short-circuits, and escape hatches.

github-actions · 2026-07-31T12:37:38Z

📊 Editorial balance

Section depth, mention distribution, recommendation steering

Section depth: 12 H2 sections (mean 90 lines, median 27.5, std 201.5). Outliers: Core Infrastructure as Code Tools: 753 (27.4× median), Frequently Asked Questions: 98 (3.6× median).
Vendor / entity mentions: counted across the tables this PR adds or rewrites, coverage is symmetric — all three comparison tables (at-a-glance, pricing, license/governance) give each of the ten tools exactly one row: Pulumi · Terraform · AWS CDK · AWS CloudFormation · Azure ARM · Azure Bicep · Google Cloud Infrastructure Manager · Kubernetes YAML · Crossplane · OpenTofu. Spacelift and env0 appear once each, as a parenthetical aside below the pricing table. The new "which tool fits your situation" table is the one asymmetric case: 8 rows, Pulumi recommended in 2 (multi-cloud + one language; AI agents), and Terraform, OpenTofu, AWS CDK, AWS CloudFormation, Azure Bicep, and Crossplane recommended in 1 each.
FAQ steering: this PR adds 3 FAQ entries — 1 steers to Pulumi (the AI-coding-agents question, which names Pulumi Neo and contrasts it against HCL), and 2 are vendor-neutral (the cost question spans all ten tools plus HCP Terraform; the popularity question answers with Kubernetes and Terraform ahead of Pulumi on GitHub stars).

🚨 Outstanding in this PR

These must be resolved or refuted before merging.

[L163] content/blog/infrastructure-as-code-tools/index.md — "Spacelift and env0 price by usage-based or per-resource quotes rather than published tiers." — verdict: contradicted; framing: Claim asserts pricing is entirely usage-based/per-resource quotes "rather than published tiers," but both vendors publish named entry-level tiers with fixed st…; evidence: (escalated from pass1) Both vendors publish named tiers with starting prices, not just usage-based quotes: Spacelift lists a Free plan and "Starter plan begins at $399/month for 10 users," with only higher tiers (Business/Enterprise) using…; source: https://zoftwarehub.com/products/spacelift/pricing ; https://www.getapp.com/it-management-software/a/env0/pricing/ ; https://www.saasworthy.com/product/env0/pricing

Both vendors do publish named tiers with starting prices — Spacelift has a Free plan and a Starter plan from $399/month for 10 users; only their Business/Enterprise tiers are quote-based. As written, the sentence tells a reader not to expect published figures that are in fact published, which is the kind of detail a competitor will notice in a comparison post. Suggested rewrite that keeps your point (they aren't denominated like the table above) without the inaccuracy:
```
-Spacelift and env0, two managed CI/CD layers frequently compared against these core tools, price by usage-based or per-resource quotes rather than the tiers above; check their sales pages directly, since published figures change often and aren't always denominated the same way.
+Spacelift and env0, two managed CI/CD layers frequently compared against these core tools, publish their own tiers but denominate them differently — largely per-seat and per-worker rather than per-resource — with quote-based pricing at the top end. Check their pricing pages directly, since the figures change often and don't map cleanly onto the table above.
```
[L1489] content/blog/infrastructure-as-code-tools/index.md — "Agents working in HCL with Terraform and OpenTofu lack access to native unit testing and IDE tooling." — verdict: contradicted; evidence: Both Terraform and OpenTofu have native unit testing support: "Both Terraform (v1.6+) and OpenTofu support native unit testing using HCL-based .tftest.hcl files, no additional language or framework required." IDE tooling is also widely ava…; source: https://spacelift.io/blog/terraform-test

This one is worth getting right because it's the claim most likely to be quoted back at us. Terraform has shipped a native test framework since 1.6 (terraform test with .tftest.hcl files), OpenTofu has the same, and both have first-party language servers and editor extensions — so "lack access to native unit testing, IDE tooling" is straightforwardly false, not just strongly worded. There's a real and defensible version of this argument: the ceiling is lower, because HCL tests can't reach for a general-purpose language's assertion libraries, mocking, property-based testing, or package ecosystem. Suggested rewrite:
```
-Declarative DSL-based tools such as Terraform and OpenTofu can still be used with AI coding assistants, but agents working in HCL lack access to native unit testing, IDE tooling, and the broader software-engineering toolchain that make agent-generated changes easy to validate before they ship.
+Declarative DSL-based tools such as Terraform and OpenTofu can also be driven by AI coding assistants, and both ship a native test framework (`terraform test` / `tofu test`) plus language servers and editor extensions. What agents don't get in HCL is the depth of the surrounding toolchain — the assertion and mocking libraries, package ecosystem, and refactoring tools that come free with a general-purpose language — so validating an agent-generated change tends to take more bespoke scaffolding.
```
Two things to fold in while you're here. First, per AGENTS.md §"AI and agent positioning", the answer should read as a tradeoff rather than a deficit, and the phrase "opaque HCL diffs" at L197 leans the same way — "HCL diffs, which are harder for an agent to test against" would land the same point without the swipe. Second, this FAQ names Pulumi Neo as the example agent; a sentence acknowledging that Claude Code, Cursor, and Codex work with Pulumi through the same code-first path would match how the migration guides handle this and would make the answer more credible, not less.

⚠️ Low-confidence

Review each and resolve as appropriate — these don't block the PR.

[L5] content/blog/infrastructure-as-code-tools/index.md — "lastmod: 2026-07-31" — two frontmatter conventions to settle before merge.

lastmod shouldn't be here. AGENTS.md §"Dates: updated vs lastmod" is explicit that updated is the reader-facing field on this site and that lastmod should not be hand-stamped — the site sets enableGitInfo: true, so Hugo already derives .Lastmod from the commit date. The updated: 2026-07-31 you added on L4 is correct and is what renders as the "Updated" line; the lastmod line beside it is invisible to readers and redundant with git:
```
 updated: 2026-07-31
-lastmod: 2026-07-31
```
The byline change deserves a human sign-off. L9-10 replaces asaf-ashirov and isaac-harris with pulumi-content-team. That's a defensible call for an evergreen listicle that's been substantially rewritten, but removing named authors from a published post isn't something a review can approve on their behalf — please confirm with them (or with whoever owns this page) before merging.
[L37] content/blog/infrastructure-as-code-tools/index.md — "As of July 2026, Pulumi's package registry lists 308 unique packages, comprising first-party providers, bridged Terraform providers, and community components." — verdict: unverifiable; evidence: No authoritative Pulumi source publishes an exact current count of registry packages. The registry page and community-packages list confirm many packages exist and continue to grow, but no page states "308" as a total; the closest historic…; source: WebSearch ran query "Pulumi package registry number of packages 2026"; top results (pulumi.com/registry/, pulumi.com/blog/introducing-pulumi-registry/) didn't state a current total count; intuition: A precise, non-round number like "308" for a live, constantly-changing registry count is hard to verify and looks like…

Author question: where did "380 package entries spanning 308 unique packages" come from? No Pulumi page publishes either total, so the figures read as a hand count of the registry index — which is fine, but it means nobody can check them later and they'll silently rot. Two precise numbers in the first paragraph of an SEO-facing post is also more exactness than the sentence needs, since the point is "coverage is broad." Consider either citing how they were derived or rounding to a number you're happy to defend for a year:
```
-As of July 2026, [Pulumi's own package registry](https://www.pulumi.com/registry/packages.md) lists 380 package entries spanning 308 unique packages (first-party providers, bridged Terraform providers, and community components), illustrating how far multi-cloud coverage has expanded across the ecosystem.
+As of July 2026, [Pulumi's own package registry](https://www.pulumi.com/registry/packages.md) lists more than 300 packages — first-party providers, bridged Terraform providers, and community components — illustrating how far multi-cloud coverage has expanded across the ecosystem.
```
Note this also needs to stay consistent with the "170+ providers" figure used at L131, L212, and L191 — right now the post cites provider coverage three different ways.
[L192] content/blog/infrastructure-as-code-tools/index.md — "OpenTofu offers drop-in HCL compatibility." — verdict: framing-drift; framing: overclaim — article body says "high, but not full, compatibility... with some divergence emerging"; summary table claims unqualified "drop-in HCL compatibility"; evidence: The same blog post's own detailed OpenTofu section (line 886) qualifies this claim, stating OpenTofu provides "high, but not full, compatibility with Terraform" and notes "some divergence emerging as the projects evolve independently" (lin…; source: content/blog/infrastructure-as-code-tools/index.md (self-referential, lines 886, 891 vs. 192); intuition: Absolute "drop-in" framing conflicts with the article's own more careful/hedged language elsewhere in the same post.

The post contradicts itself: the new "which tool fits your situation" table promises unqualified "drop-in HCL compatibility," while the OpenTofu section at L886 says OpenTofu offers "high, but not full, compatibility with Terraform" and notes "some divergence emerging as the projects evolve independently." L886 is the accurate version (compatibility is complete for Terraform pre-1.6 state and providers, and diverges after that), so the table is what should move:
```
-| You need an open-source, community-governed Terraform-compatible tool | OpenTofu | MPL-2.0, Linux Foundation governance, drop-in HCL compatibility |
+| You need an open-source, community-governed Terraform-compatible tool | OpenTofu | MPL-2.0, Linux Foundation governance, high HCL compatibility |
```
A skimmer who reads only the table would migrate expecting no rewrites, which is the worst place for this post to be optimistic.
[L212] content/blog/infrastructure-as-code-tools/index.md — "Pulumi provides native cloud provider SDKs enabling same-day feature access across 170+ providers." — verdict: framing-drift; framing: overclaim — source scopes "same-day support for new features" to native providers (AWS/Azure/GCP/K8s), claim applies it across "170+ providers" in the registry; evidence: Same-day feature access is a documented characteristic specifically of Pulumi's Native Providers (AWS Native, Azure Native, Google Cloud Native), generated automatically from cloud provider APIs: "Native providers are built directly and au…; source: https://www.pulumi.com/blog/pulumiup-native-providers/ and https://www.pulumi.com/blog/any-terraform-provider/

The bump from "150+" to "170+" is fine, but this line staples two separate facts together. Same-day feature access is a property of the native providers (AWS Native, Azure Native, Google Cloud Native, Kubernetes), which are generated from the cloud vendors' own API specs. The 170+ figure counts the whole registry, most of which is bridged Terraform providers that track upstream releases rather than the cloud APIs directly. As written, the bullet promises same-day coverage across all 170+:
```
-- **Multi-cloud flexibility**: Native cloud provider SDKs with same-day feature access across [170+ providers](/registry/)
+- **Multi-cloud flexibility**: [170+ providers](/registry/), with native AWS, Azure, Google Cloud, and Kubernetes providers offering same-day access to new cloud features
```
The same conflation appears in the "why" cell at L191 ("Real programming languages ... and 170+ providers") — that one is fine as-is, since it doesn't attach the same-day promise.
[L487-490] content/blog/infrastructure-as-code-tools/index.md — "Yarin Miran, Senior Software Engineer at Wiz, said: 'We use Pulumi widely at Wiz. It enabled our product to support multi-cloud and to scale quickly - scaling…" — verdict: unverifiable; evidence: The pre-fetched body of https://www.pulumi.com/case-studies/wiz/ is truncated to generic site navigation/header content and does not include the quoted text attributed to Yarin Miran. Cannot confirm the quote appears on the cited page from…; source: https://www.pulumi.com/case-studies/wiz/

Author question: can you confirm this quote is verbatim from the Wiz case study, and that Yarin Miran's title is still accurate? Fetching https://www.pulumi.com/case-studies/wiz/ returned only the page's navigation shell, so the quote couldn't be matched against the source here — this is a limit of the check, not evidence the quote is wrong. Two of the three pull-quotes this PR adds (see [L1285-1288] and [L1315-1318]) are in the same position. Since attributed quotes carry a person's name, they're worth confirming character-for-character against the live page before merge. One nit visible in the text itself: the quote uses a hyphen where the sentence wants an em dash — "scale quickly - scaling and driving" → "scale quickly — scaling and driving" — assuming the source page does too.
[L849] content/blog/infrastructure-as-code-tools/index.md — "Latest stable release: v2.3.4 (2026-07-23)" — verdict: unverifiable; evidence: No search result identifies which tool "v2.3.4 (2026-07-23)" refers to, nor confirms such a release exists; Pulumi's own CLI version tracked by Wikipedia is 3.243.0 (May 22, 2026), not v2.3.4, so this figure does not correspond to Pulumi i…; source: WebSearch ran query "v2.3.4 latest stable release 2026-07-23"; top results didn't address the claim; intuition: Suspiciously precise version/date pair with no identifiable tool name attached; smells like a fabricated or template-fi…

For context the check was missing: this line is the ### 9. Crossplane section header, so v2.3.4 is a Crossplane version claim (it also appears in the at-a-glance table at L141 and the governance table at L177). It's the one version in the refresh that no source confirmed either way — every other tool's version and date was matched to its upstream release, and Bicep's was initially flagged wrong but confirmed correct on dispute (see 🚨 → ✅ history above). Author question: please re-check v2.3.4 / 2026-07-23 against https://github.com/crossplane/crossplane/releases and confirm the ~11.9k star count while you're there.
[L1285-1288] content/blog/infrastructure-as-code-tools/index.md — "Joe Duffy, Founder & CEO of Pulumi, said: 'In under a year, AI has completely reshaped how we build applications. With Pulumi Neo, platform engineering is now…" — verdict: unverifiable; evidence: The pre-fetched body is only the generic PR Newswire navigation/boilerplate content; the truncated body does not contain the actual press release text or Joe Duffy's quote, so the exact quote cannot be confirmed from this fetch.; source: https://www.prnewswire.com/news-releases/introducing-pulumi-neo-the-industrys-first-ai-powered-platform-engineer-302556718.html

The press release itself was confirmed — "Introducing Pulumi Neo, the Industry's First AI-Powered Platform Engineer" (PR Newswire, Sept 16, 2025) does carry a Joe Duffy quote opening "In under a year, AI has completely reshaped how we build applications." The exact wording of the rest couldn't be matched, because the fetch returned only PR Newswire's boilerplate.

What I'd change regardless of the wording check is the ellipsis. "platform engineering is now catching up. ... Things that used to take weeks" elides material from the middle of a CEO quote, and the prose linter flags the ellipsis separately (see the style findings for L1285). Quoting the two sentences you actually want, without the bridge, reads better and is safer:
```
-> "In under a year, AI has completely reshaped how we build applications. With Pulumi Neo, platform engineering is now catching up. ... Things that used to take weeks can now confidently be done in minutes."
+> "In under a year, AI has completely reshaped how we build applications. With Pulumi Neo, platform engineering is now catching up."
```
Also worth noting the quote is from September 2025, which is fine for a 2026 refresh but is old enough that the sentence introducing it shouldn't imply it's new.
[L1311] content/blog/infrastructure-as-code-tools/index.md — "Atlassian Bitbucket's rapid 2-day migration was enabled by existing Python expertise on the team, well-structured Terraform code, and straightforward infrastru…" — verdict: framing-drift; framing: source credits the automatic conversion tool for the fast timeline; claim substitutes unsupported reasons ("well-structured Terraform code, straightforward inf…; evidence: The case study confirms the 2-day timeline and Python-familiarity rationale: "Because all the developers on Bitbucket write Python, they can use all the skills they already have" and "I spent one day using the automatic conversion tool...…; source: https://www.pulumi.com/case-studies/atlassian/

This is on a line the PR doesn't touch, so it's not something the refresh introduced — but it's adjacent to the Starburst quote you're adding, so it's cheap to fix in the same pass. The case study credits the two-day timeline to the automatic conversion tool plus the team's existing Python fluency ("Because all the developers on Bitbucket write Python, they can use all the skills they already have"). The Python half is supported; "well-structured Terraform code" and "straightforward infrastructure" aren't in the source at all. Dropping the two unsupported reasons keeps the sentence true to the case study.
[L1311] content/blog/infrastructure-as-code-tools/index.md — "Atlassian Bitbucket completed a full migration from Terraform to Pulumi in 2 days using automated conversion tools." — verdict: framing-drift; framing: overclaim — source describes converting a single pipeline/project in two days; claim broadens this to "a full migration" of Bitbucket's Terraform usage to Pulu…; evidence: The Pulumi case study quotes Mike: "I spent one day using the automatic conversion tool and then just spent some time making sure everything looked right... In two days, I was all set." This describes converting one specific CI/CD pipeline…; source: https://www.pulumi.com/case-studies/atlassian/

Same untouched line, second issue: "a full migration" overstates what the case study describes. Mike's "In two days, I was all set" refers to converting one CI/CD pipeline, not all of Bitbucket's Terraform. Since a "2 days" datapoint in a migration section is exactly what a reader will anchor their own estimate to, scoping it is worth doing:
```
-- **Atlassian Bitbucket**: Completed full migration from Terraform to Pulumi's Python-based IaC in just 2 days using migration tools
+- **Atlassian Bitbucket**: Converted a Terraform-managed CI/CD pipeline to Pulumi's Python-based IaC in 2 days, using the automatic conversion tool and the team's existing Python experience
```
Note the same "2 days" framing recurs at L1460, where it was ✅ verified because that line reads more narrowly — worth making the two consistent.
[L1315-1318] content/blog/infrastructure-as-code-tools/index.md — "Matt Stephenson, Senior Principal Software Engineer at Starburst, said: 'When we did it with Terraform, it took two weeks to do [infrastructure deployments]. N…" — verdict: unverifiable; evidence: The pre-fetched body of the cited case-study URL is truncated to generic site navigation/header content and does not contain the quoted passage attributed to Matt Stephenson, so the specific quote could not be confirmed against the source.; source: https://www.pulumi.com/case-studies/starburst/

Author question: please confirm this quote verbatim against the live case study. The fetch returned only the page's navigation shell, so the wording couldn't be matched — though the underlying numbers are independently confirmed: the Starburst case study does report the two-weeks-to-hours change and a "112x faster deployment time" (verified separately at L1461).

Two things to check in the quote as written. The bracketed insertion "[infrastructure deployments]" should match whatever the source actually elides, and the prose linter flags "So that's how much..." for starting a sentence with "So" (see style findings, L1315) — that's a fair flag in general but shouldn't be applied here, since editing a direct quote to satisfy a style rule is worse than the nag. Leave the quote alone and let that one go.
[L1462] content/blog/infrastructure-as-code-tools/index.md — "Multiple organizations report 80-90% deployment time reductions after moving to programming language-based IaC approaches." — verdict: framing-drift; framing: Claim compresses varied figures (80% Unity, 99% Starburst, 70% SANS) into a uniform "80-90%" range attributed to "multiple organizations," but at least one cit…; evidence: The blog itself states "Organizations like Unity, Snowflake, and Starburst have reported significant productivity improvements (80-90% deployment time reductions)," but Pulumi's own case studies show Unity reported "80% reduction in deploy…; source: https://www.pulumi.com/case-studies/starburst/ and https://www.pulumi.com/case-studies/unity/; intuition: The specific 80-90% range doesn't match the actual spread of reported figures (70%-99%) across Pulumi's own case studie…

On an untouched line, and the same "80-90%" figure flagged as ❌ contradicted at [L497] (see 💡 Pre-existing) — the two lines should be fixed together. The real numbers from Pulumi's own case studies are Unity at 80%, SANS at 70%, and Starburst at 99% (112x), so "80-90%" both narrows the range and excludes two of the three companies actually cited. The honest version is stronger than the invented range, since it includes the 99%:
```
-Multiple organizations report 80-90% deployment time reductions after moving to programming language-based IaC approaches.
+Organizations moving to programming-language-based IaC report deployment-time reductions ranging from roughly 70% to 99% — Unity at 80%, SANS at 70%, and Starburst at 99% (two weeks down to about three hours).
```
Because this refresh is otherwise careful to cite versions and prices to the day, a range that doesn't match the linked case studies is the kind of thing that undermines the rest of the numbers.

Style findings

Found by pattern-based linting; Findings may be false positives.

line 41: [style] difficulty qualifier — Avoid difficulty qualifier 'simple' -- it judges difficulty for the reader (STYLE-GUIDE.md §Inclusive Language).
line 41: [style] difficulty qualifier — Avoid difficulty qualifier 'just' -- it judges difficulty for the reader (STYLE-GUIDE.md §Inclusive Language).
line 148: [style] difficulty qualifier — Avoid difficulty qualifier 'just' -- it judges difficulty for the reader (STYLE-GUIDE.md §Inclusive Language).
line 1285: [style] weasel word — 'completely' is a weasel word!
line 1285: [style] punctuation — In general, don't use an ellipsis.
line 1315: [style] filler — Don't start a sentence with 'So '.
line 1489: [style] difficulty qualifier — Avoid difficulty qualifier 'easy' -- it judges difficulty for the reader (STYLE-GUIDE.md §Inclusive Language).

📋 Triaged verifier findings

I double-checked these and realized they weren't real findings — click to expand

[L7] content/blog/infrastructure-as-code-tools/index.md — "Pulumi, Terraform, OpenTofu, AWS CDK, AWS CloudFormation, Azure ARM, Azure Bicep, Google Cloud Infrastructure Manager, Kubernetes YAML, and Crossplane are comp…" — verdict: unverifiable; framing: Cited source is unrelated to the claim's content (a package index vs. a 2026 tool-comparison analysis); page was fetched successfully but does not address the…; evidence: The cited URL is just the Pulumi Registry packages index listing (an alphabetical list of provider packages like AWS, Azure, etc.) and does not contain any content comparing Pulumi, Terraform, OpenTofu, AWS CDK, CloudFormation, ARM, Bicep,…; source: https://www.pulumi.com/registry/packages.md

Mis-sourced: the registry link in that paragraph supports the package-count figure later in the sentence, not the list of ten tools, but the check attached it to the tool list and then found the page unrelated. The list of ten tools is the article's own statement of scope — which tools it chose to cover — and isn't an external claim needing a citation. The package count that link does support is handled separately at [L37].
[L558] content/blog/infrastructure-as-code-tools/index.md — "Latest stable release: v2.262.2 (2026-07-29)" — verdict: contradicted; evidence: No tool in this space has version v2.262.2. Terraform's latest stable is reported as "Terraform 1.15.x (1.15.8 as of July 2026)" and Pulumi's latest releases are v3.254.0/v3.255.0 (per GitHub releases page, July 2026). The claimed version…; source: WebSearch ran query "Terraform latest stable release v2.262.2" and "Pulumi CLI latest release version July 2026"; top results show Terraform at 1.15.x and Pulumi at v3.254.0/v3.255.0, not v2.262.2

Spurious: the line sits under ### 3. AWS Cloud Development Kit (CDK), so v2.262.2 is an AWS CDK version — but the check read it without its section heading and compared it against Terraform and Pulumi release numbering, which of course didn't match. v2.262.2 (2026-07-29) is correct for aws/aws-cdk and was independently confirmed twice elsewhere in this review (L135 and L173).
[L883] content/blog/infrastructure-as-code-tools/index.md — "Latest stable release: v1.12.5 (2026-07-21)" — verdict: contradicted; evidence: Multiple sources confirm Terraform's latest stable release as of July 2026 is in the v1.15.x series (e.g. "The latest stable release is Terraform 1.15.x (1.15.8 as of July 2026)"), not v1.12.5, which is an older, superseded minor version l…; source: https://spacelift.io/blog/terraform-version; intuition: Claimed version v1.12.5 is several minor versions behind the actual current release (~v1.15.8) as of the claim's own da…

Spurious: same context problem — this line is under ### 10. OpenTofu, so v1.12.5 is an OpenTofu version, not Terraform's. Because OpenTofu is a Terraform fork with parallel v1.x numbering, comparing it against Terraform's v1.15.8 makes it look several minors stale. v1.12.5 (2026-07-21) is confirmed correct for opentofu/opentofu at L142 and L180.

💡 Pre-existing issues in touched files (optional)

[L497] content/blog/infrastructure-as-code-tools/index.md — "Unity, Snowflake, and Starburst reported significant productivity improvements, specifically 80-90% deployment time reductions, when adopting programming langu…" — verdict: contradicted; framing: Claim invents a uniform "80-90%" range and attributes it to all three companies; actual reported figures are a 112x/99% reduction for Starburst, an unquantifie…; evidence: Pulumi's own case studies report far more dramatic and differently-framed figures than "80-90% deployment time reductions": Starburst reports a "112x faster deployment time" (a 99% reduction, two weeks to three hours), and Snowflake report…; source: https://www.pulumi.com/case-studies/starburst/ ; https://www.pulumi.com/case-studies/snowflake/ ; https://www.pulumi.com/case-studies/unity/; intuition: Suspiciously round shared range (80-90%) attributed uniformly to three different companies whose actual reported number…

Pre-existing: this sentence is on an untouched line — the PR only adds the Wiz pull-quote a few lines above it. The "80-90%" range doesn't match any of the three linked case studies (Starburst reports 99% / 112x, Unity 80%, and Snowflake's figure isn't quantified that way), and Snowflake in particular is named without a matching number. Since this refresh is explicitly about currency and it's four lines from a block you're already editing, it's worth correcting in this PR rather than leaving for later. The same range recurs at [L1462]; fix both together, and see that bullet for suggested wording.
[L565] content/blog/infrastructure-as-code-tools/index.md — "AWS CDK supports TypeScript, Python, Java, C#, and JavaScript with full IDE integration, and Go is available in Developer Preview." — verdict: contradicted; framing: shifted — claim states Go is currently "Developer Preview" (true circa 2021-2022) but current AWS docs say Go "is considered stable" and fully supported, so th…; evidence: (escalated from pass1) Current AWS CDK documentation states Go has graduated from Developer Preview: "Go is a fully-supported client language for the AWS Cloud Development Kit (AWS CDK) and is considered stable." The v2 languages page also…; source: https://docs.aws.amazon.com/cdk/v2/guide/work-with-cdk-go.html; intuition: Go's Developer Preview status for AWS CDK ended years ago; this claim appears to use stale information.

Pre-existing: on an untouched line in the AWS CDK section — the PR adds the License: and Latest stable release: lines just above it but doesn't touch the feature list. Still, it's stale in a way that matters for a post whose headline promise is "for 2026": AWS's own docs now say Go "is a fully-supported client language ... and is considered stable," and the at-a-glance table you rewrote at L135 already lists Go without a preview caveat, so the page currently contradicts itself. One-line fix while you're in the section:
```
-- **Multi-language support**: TypeScript, Python, Java, C#, and JavaScript with full IDE integration (Go available in Developer Preview)
+- **Multi-language support**: TypeScript, Python, Java, C#, JavaScript, and Go with full IDE integration
```

✅ Resolved since last review

[L138] content/blog/infrastructure-as-code-tools/index.md — "| Azure Bicep | Bicep DSL (compiles to ARM JSON) | Azure only | MIT | v0.46.1 (2026-07-30) | Azure deployments requiring improved readability…" — concede: @CamSoper disputed this finding ("The current version of Bicep is v0.46.1 (from your own link!)"). Re-checked Azure/bicep via gh api repos/Azure/bicep/releases/latest: tag v0.46.1, published 2026-07-30T23:35:23Z, prerelease: false. The table's version and date are correct — the original finding's source evidently returned a stale snapshot of the releases page.
[L176] content/blog/infrastructure-as-code-tools/index.md — "| Azure Bicep | MIT | Microsoft | v0.46.1 | 2026-07-30 | ~3.6k |" — concede: same dispute and re-check as [L138] — v0.46.1 (2026-07-30) is genuinely the latest Bicep release; the license, governing body, and star count were already correct.
[L697] content/blog/infrastructure-as-code-tools/index.md — "Latest stable release: v0.46.1 (2026-07-30)" — concede: same dispute and re-check as [L138] — the Azure Bicep section header's version is correct.

📜 Review history

2026-07-31T12:30:09Z — Verified most of the refresh's new version, license, and pricing figures against upstream sources; 5 blockers remain (Azure Bicep listed as v0.46.1 in three places when the latest is v0.45.15, a wrong Spacelift/env0 pricing claim, and a false "HCL lacks native unit testing" assertion in the new AI-agents FAQ), plus unconfirmed Crossplane version, three unverified pull-quotes, and a lastmod/byline frontmatter question. (7e049b8)
2026-08-04T00:52:21Z — @CamSoper disputed the Azure Bicep version finding ("The current version of Bicep is v0.46.1 (from your own link!)"). Re-checked Azure/bicep via gh api repos/Azure/bicep/releases/latest: tag v0.46.1, published 2026-07-30T23:35:23Z, prerelease: false — confirmed correct, not a stale/prerelease artifact. Conceded all three occurrences (L138, L176, L697) to ✅ Resolved. No new commits since the last review (head still 7e049b8); 2 outstanding findings unchanged (Spacelift/env0 pricing at L163, HCL unit-testing claim at L1489).

Refresh this review — comment @claude #update-review. Say what you fixed, or which finding you dispute and why; both work in the same mention.
Ask for anything else — comment @claude with no hashtag (questions, one-off fixes). Leaves this review untouched.

Important

Please don't hide, resolve, or delete this comment! It breaks things!

📖 How pre-merge review works — the full lifecycle, short-circuits, and escape hatches.

pulumi-bot · 2026-07-31T12:38:27Z

Your site preview for commit 7e049b8 is ready! 🎉

http://www-testing-pulumi-docs-origin-pr-20609-7e049b86.s3-website.us-west-2.amazonaws.com

Changed pages:

/blog/infrastructure-as-code-tools/

CamSoper · 2026-08-04T00:48:02Z

@claude The current version of Bicep is v0.46.1 (from your own link!). #update-review

@workprentice please address the rest of the pre-merge feedback, including the low-confidence and style items.

github-actions · 2026-08-04T01:06:18Z

🤖 Review updated on @CamSoper's request.

workprentice Bot temporarily deployed to production July 31, 2026 12:26 Inactive

workprentice Bot temporarily deployed to testing July 31, 2026 12:26 Inactive

github-actions Bot added review:triaging Claude Triage is currently classifying the PR domain:blog PR touches blog posts or customer stories review:in-progress Claude review is currently running and removed review:triaging Claude Triage is currently classifying the PR labels Jul 31, 2026

github-actions Bot added review:outstanding-issues Claude review completed; outstanding has author-actionable findings and removed review:in-progress Claude review is currently running labels Jul 31, 2026

github-actions Bot added review:in-progress Claude review is currently running and removed review:outstanding-issues Claude review completed; outstanding has author-actionable findings labels Aug 4, 2026

github-actions Bot added review:outstanding-issues Claude review completed; outstanding has author-actionable findings and removed review:in-progress Claude review is currently running labels Aug 4, 2026

Navigation Menu

Uh oh!

Refresh: Best Infrastructure as Code (IaC) Tools for 2026 — tables, versions, quotes, FAQs - #20609

Refresh: Best Infrastructure as Code (IaC) Tools for 2026 — tables, versions, quotes, FAQs#20609
workprentice[bot] wants to merge 1 commit into
masterfrom
refresh/iac-tools-listicle-2026-07

workprentice Bot commented Jul 31, 2026

Uh oh!

github-actions Bot commented Jul 31, 2026 •

edited by pulumi-bot

Loading

Uh oh!

github-actions Bot commented Jul 31, 2026 •

edited by pulumi-bot

Loading

Uh oh!

pulumi-bot commented Jul 31, 2026

Uh oh!

CamSoper commented Aug 4, 2026

Uh oh!

github-actions Bot commented Aug 4, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

workprentice Bot commented Jul 31, 2026

Summary

Why: mention vs. citation gap on Perplexity

Competitor teardown

What changed

Schema notes (verified against layouts/partials/schema/, not assumed)

Data sources

Uh oh!

github-actions Bot commented Jul 31, 2026 • edited by pulumi-bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pre-merge Review — Last updated 2026-08-04T00:52:21Z

🔍 Verification trail

Uh oh!

github-actions Bot commented Jul 31, 2026 • edited by pulumi-bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

📊 Editorial balance

🚨 Outstanding in this PR

⚠️ Low-confidence

Style findings

📋 Triaged verifier findings

💡 Pre-existing issues in touched files (optional)

✅ Resolved since last review

📜 Review history

Uh oh!

pulumi-bot commented Jul 31, 2026

Uh oh!

CamSoper commented Aug 4, 2026

Uh oh!

github-actions Bot commented Aug 4, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

github-actions Bot commented Jul 31, 2026 •

edited by pulumi-bot

Loading

github-actions Bot commented Jul 31, 2026 •

edited by pulumi-bot

Loading