Pentest report example
A Cloudwrit pentest in the “literary form” we defend. Claim, evidence, proof-of-concept, remediation, retest window.
The specimens on this page are the shape of what we ship. Some are excerpts from real engagements (sanitized, with consent). Others are fictional but representative, built on our Cloudwrit example so the voice stays consistent. Every sample is labeled.
A Cloudwrit pentest in the “literary form” we defend. Claim, evidence, proof-of-concept, remediation, retest window.
STRIDE output for Cloudwrit. Attack-path diagram plus mitigation-to-sprint mapping.
Severity-tagged checklist with inline Terraform snippets for AWS and GCP.
Top 10 performance failures and top 8 accessibility failures with fix code.
Anonymized engagement spec: auth, billing, admin, RBAC, analytics. What we ship in 4 weeks.
Golden dataset plus scorer results plus regression dashboard for a doc-summarization feature. JSON export.
Token economics calculator. Inputs, scenarios, caching ROI. Copy-friendly values.
5 prompts versioned v1 through v3, with eval scores per version.
Dashboard view of the raw export from our analysis. JSON export.
Pillar and cluster visualization for a fictional SaaS. CSV export.
Full positioning output: context, ICP, UVP, messaging hierarchy, anti-messaging.
5-email sequence targeting CTO ICP. Personalization inputs, expected response rates.
The 15 to 20 page Cloudwrit handbook subset. Companion to the essay.
Samples marked Fictional are built on our Cloudwrit example client. The shape of the document, the tone, the structure - those are real. The numbers and companies are invented. Samples marked Real are from genuine work with the data anonymized. Template means the sample is the same artifact we use internally - you can copy it.
If you want the real version of something that only appears here as fictional, start a conversation - we are usually able to share anonymized excerpts from prior engagements under NDA.