Skip to content
Back to skills

Report Writing

ASecurity

Red team and penetration test report writing methodology. Covers executive summary, technical findings format, CVSS4 scoring, remediation guidance, and evidence documentation standards.

  • 5 stars
  • 0 votes
  • 0 copies
  • 1 view
  • Added September 25, 2026
ai-agentsgosqltestingapidatabasedocumentation

Works with

  • terminal
  • api

Security analysis

A100/100

Scanned September 25, 2026

npx -y skills add sunilgentyala/OmniRed --skill report-writing --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Report Writing?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Report Writing
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/sunilgentyala-report-writing/badge)](https://www.skillsdirectory.com/skills/sunilgentyala-report-writing)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: report-writing
description: Red team and penetration test report writing methodology. Covers executive summary, technical findings format, CVSS4 scoring, remediation guidance, and evidence documentation standards.
version: 1.0.0
license: Apache-2.0
---

# Penetration Test Report Writing

## Report Structure

```
1. Cover Page
   - Engagement title, target, dates, report date
   - Classification: CONFIDENTIAL

2. Executive Summary (1-2 pages)
   - Overall risk rating
   - Critical findings summary (non-technical)
   - Business impact statement
   - Top 3 remediation priorities

3. Scope and Methodology
   - In-scope assets, IP ranges, domains
   - Testing approach (black/grey/white box)
   - Tools used
   - Testing dates and testers

4. Risk Rating Matrix
   - CVSS v4.0 base scores
   - Environmental modifiers

5. Technical Findings (one section per finding)

6. Remediation Roadmap
   - Priority order
   - Estimated effort
   - Quick wins vs. strategic fixes

7. Appendices
   - Raw scan output
   - Payload lists
   - Tool configurations
```

## Finding Format (per vulnerability)

```markdown
## FINDING-001: [Vulnerability Name]

**Severity:** Critical | High | Medium | Low | Informational
**CVSS v4.0 Score:** 9.3 (AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N)
**CWE:** CWE-89 (SQL Injection)
**OWASP:** A03:2021 — Injection
**MITRE ATT&CK:** T1190 — Exploit Public-Facing Application

### Description
[Plain English description of the vulnerability — what it is and why it exists]

### Business Impact
[What an attacker can do if they exploit this — in business terms, not technical terms]

### Evidence
**Request:**
```http
POST /api/login HTTP/1.1
Host: target.com
Content-Type: application/json

{"username":"admin' OR '1'='1","password":"x"}
```

**Response:**
```http
HTTP/1.1 200 OK
{"token":"eyJ...","role":"admin"}
```

**Screenshot:** [embed screenshot]

### Steps to Reproduce
1. Navigate to https://target.com/api/login
2. Submit the above request
3. Observe admin authentication without valid credentials

### Remediation
- **Short-term (1 week):** Implement parameterised queries using prepared statements
- **Long-term (1 month):** Migrate to ORM with built-in SQLi protection; add WAF rule
- **References:**
  - OWASP SQL Injection Prevention Cheat Sheet
  - CWE-89 remediation guidance
```

## CVSS v4.0 Quick Reference

```
Metric          Values
AV (Vector)     N=Network, A=Adjacent, L=Local, P=Physical
AC (Complexity) L=Low, H=High
AT (Reqs)       N=None, P=Present
PR (Privs)      N=None, L=Low, H=High
UI (Interaction) N=None, P=Passive, A=Active
VC/VI/VA        H=High, L=Low, N=None (Confidentiality/Integrity/Availability)
SC/SI/SA        H=High, L=Low, N=None (Subsequent system C/I/A)
```

## Executive Summary Writing Guide

- Lead with the most critical finding in plain business language
- Quantify: "An attacker could access all 50,000 customer records"
- Avoid technical jargon: "SQL injection" → "A flaw that lets attackers read any data in the database"
- State the risk rating clearly: "We rate the overall risk as CRITICAL"
- End with 3 actionable priorities

## Evidence Standards

- All screenshots must show: URL/host, timestamp, your test annotation
- HTTP requests/responses: full headers + body
- Command output: full terminal with prompt showing hostname
- For blind vulnerabilities: out-of-band callback logs (Burp Collaborator)
- Video recordings for complex multi-step exploits

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.