OSCP Report Writing: Evidence Collection and Screenshots
1. Why Evidence Matters
The Evidence Requirement
In penetration testing, a claim without evidence is just an opinion. Every finding in your report must be supported by concrete evidence that proves the vulnerability exists and is exploitable.
For the OSCP exam specifically:
- Screenshots are mandatory for each finding
- You must show proof of exploitation (not just detection)
- You must show flag capture (user.txt and root.txt/proof.txt)
- Evidence must be clear, readable, and annotated
Evidence Types
| Evidence Type | Description | When to Use | OSCP Relevance |
|---|---|---|---|
| Screenshots | Visual proof of vulnerability and exploitation | All findings | Mandatory |
| Command Output | Terminal output showing exploit execution | Technical findings | Mandatory |
| HTTP Requests/Responses | Burp Suite / curl output showing web vulnerabilities | Web vulnerabilities | High |
| Code Snippets | Vulnerable source code (if available) | Source code review | Medium |
| Network Captures | Wireshark / tcpdump output | Network vulnerabilities | Medium |
| File Hashes | SHA256 of captured files | File analysis | Low |
| Log Entries | System/application logs | Authentication issues | Medium |
| Shell Access | Proof of shell / user access | Post-exploitation | Mandatory |
2. Screenshot Best Practices
What Makes a Good Screenshot
| Do | Don’t |
|---|---|
| Capture the full context | Crop out important information |
| Annotate with arrows and highlights | Leave screenshots unmarked |
| Include timestamps | Remove identifying information |
| Show both request and response | Show only partial output |
| Use consistent naming convention | Use generic names like “screenshot1.png” |
| Redact sensitive client data | Include real credentials in evidence |
| Show the prompt (user@host) | Hide the command prompt |
| Show the full terminal output | Cut off the bottom of the output |
Screenshot Naming Convention
{finding-id}_{description}_{timestamp}.png
Examples:
VULN-001_sqli-login-bypass_20260815.png
VULN-001_sqli-data-extraction_20260815.png
VULN-002_priv-esc-kernel_20260816.png
VULN-002_root-flag-capture_20260816.png
Screenshot Annotation Techniques
- Arrows: Point to the specific part of the screenshot that matters
- Boxes/Rectangles: Highlight the area of interest
- Text Labels: Add explanatory text to clarify what the viewer should see
- Redaction: Black out sensitive information not relevant to the finding
Annotation Tools
| Tool | Platform | Key Features |
|---|---|---|
| GIMP | Cross-platform | Full-featured image editing |
| Paint.NET | Windows | Lightweight, good for annotations |
| gnome-screenshot | Linux | Built-in screenshot tool |
| Flameshot | Linux | Advanced screenshot with annotation |
| Snipping Tool | Windows | Built-in, basic annotation |
| Burp Suite | Cross-platform | Built-in screenshot for HTTP requests |
3. Evidence Collection Checklist
During the Engagement
□ Take screenshot BEFORE running the exploit (showing the target state)
□ Take screenshot OF the exploit command/payload
□ Take screenshot OF the exploit result/output
□ Take screenshot OF the post-exploitation state
□ Take screenshot OF the flag capture (user.txt / root.txt)
□ Copy command output to notes file
□ Save HTTP request/response from Burp Suite
□ Record timestamp of each activity
□ Note the affected system and component
□ Hash evidence files (for professional engagements)
OSCP-Specific Evidence Requirements
For each machine you compromise, you MUST provide:
-
Initial Access Evidence
- Screenshot of the exploitation command
- Screenshot of the successful shell/reverse shell
- Screenshot showing the initial user context (whoami, id)
-
User Flag Evidence
- Screenshot of reading user.txt
- Screenshot showing the flag content
-
Privilege Escalation Evidence
- Screenshot of the privilege escalation method
- Screenshot showing the escalated user context
-
Root Flag Evidence
- Screenshot of reading root.txt (Linux) or proof.txt (Windows)
- Screenshot showing the flag content
4. Evidence Collection by Vulnerability Type
Web Application Vulnerabilities
What to capture:
- Burp Suite request and response
- The vulnerable parameter highlighted
- The payload used
- The response showing successful exploitation
- Browser screenshot showing the impact
Example:
Screenshot 1: Burp Suite request showing the SQL injection payload
Screenshot 2: Burp Suite response showing the extracted data
Screenshot 3: Browser showing the impact (e.g., authentication bypass)
Service Exploitation
What to capture:
- nmap scan output showing the vulnerable service
- The exploitation command (Metasploit, custom script, etc.)
- The successful exploitation output
- The shell prompt showing access
Example:
Screenshot 1: nmap output showing Apache Struts 2.5.30 on port 80
Screenshot 2: Metasploit console showing exploit configuration
Screenshot 3: Metasploit console showing successful exploitation
Screenshot 4: Shell prompt showing www-data@target:~$
Privilege Escalation (Linux)
What to capture:
- Initial user context (whoami, id)
- Privilege escalation enumeration output (LinPEAS)
- The privilege escalation exploit command
- The root shell prompt
- Reading root.txt
Example:
Screenshot 1: whoami showing initial user (e.g., www-data)
Screenshot 2: LinPEAS output showing the vulnerability
Screenshot 3: Kernel exploit compilation and execution
Screenshot 4: whoami showing root
Screenshot 5: cat /root/root.txt showing the flag
Privilege Escalation (Windows)
What to capture:
- Initial user context (whoami, whoami /priv)
- Privilege escalation enumeration output (WinPEAS)
- The privilege escalation exploit command
- The elevated shell prompt
- Reading proof.txt
Example:
Screenshot 1: whoami showing initial user (e.g., NT AUTHORITY\NETWORK SERVICE)
Screenshot 2: WinPEAS output showing the vulnerability
Screenshot 3: Exploit execution (e.g., abuse of AlwaysInstallElevated)
Screenshot 4: whoami showing NT AUTHORITY\SYSTEM
Screenshot 5: type C:\Users\Administrator\Desktop\proof.txt showing the flag
5. Evidence Collection Tools
Screenshot Tools
| Tool | Platform | Key Features | Best For |
|---|---|---|---|
| gnome-screenshot | Linux | Region selection, delay, clipboard | Quick Linux screenshots |
| Flameshot | Linux | Annotation, upload, effects | Annotated screenshots |
| Scrot | Linux | Command-line, delay, selection | Scripted screenshots |
| Snipping Tool | Windows | Built-in, basic annotation | Quick Windows screenshots |
| Snip & Sketch | Windows 10/11 | Built-in, annotation, search | Modern Windows screenshots |
| Burp Suite | Cross-platform | HTTP request/response capture | Web vulnerability evidence |
Terminal Output Capture
| Tool | Platform | Key Features | Best For |
|---|---|---|---|
| script | Linux | Records entire terminal session | Complete session recording |
| tmux | Linux | Session management, logging | Persistent terminal sessions |
| screen | Linux | Session management, logging | Persistent terminal sessions |
| tee | Linux | Redirects output to file and screen | Capturing command output |
| Get-Content | Windows | PowerShell output capture | Windows command output |
Network Capture Tools
| Tool | Platform | Key Features | Best For |
|---|---|---|---|
| Wireshark | Cross-platform | GUI packet analysis | Detailed network analysis |
| tcpdump | Linux/macOS | CLI packet capture | Lightweight capture |
| tshark | Cross-platform | CLI Wireshark | Scripted packet analysis |
File Hashing Tools
| Tool | Platform | Key Features | Best For |
|---|---|---|---|
| sha256sum | Linux | SHA256 hashing | Quick file hashing |
| hashdeep | Cross-platform | Multiple hash algorithms | Comprehensive hashing |
| certutil | Windows | Built-in hashing | Windows file hashing |
6. Evidence Storage and Organization
Directory Structure
evidence/
├── VULN-001/
│ ├── sqli-login-bypass-request.png
│ ├── sqli-login-bypass-response.png
│ ├── sqli-data-extraction.png
│ └── sqli-sqlmap-output.txt
├── VULN-002/
│ ├── priv-esc-initial-user.png
│ ├── priv-esc-linpeas-output.png
│ ├── priv-esc-exploit-execution.png
│ ├── priv-esc-root-shell.png
│ └── root-flag-capture.png
├── VULN-003/
│ └── ...
└── misc/
├── nmap-initial-scan.txt
├── network-diagram.png
└── engagement-timeline.txt
Evidence File Naming
{finding-id}_{evidence-type}_{description}.{extension}
Examples:
VULN-001_request_sqli-payload.png
VULN-001_response_data-extraction.png
VULN-002_terminal_root-shell.png
VULN-003_burp_xss-payload.png
7. Evidence Integrity
Hashing Evidence Files
For professional engagements, hash all evidence files to prove they have not been altered:
# Hash a single file
sha256sum VULN-001_sqli-login-bypass.png
# Hash all files in a directory
find evidence/ -type f -exec sha256sum {} \; > evidence_hashes.txt
# Verify hashes later
sha256sum -c evidence_hashes.txt
Chain of Custody
For legal engagements, maintain a chain of custody log:
| Date/Time | Action | Performed By | Notes |
|---|---|---|---|
| 2026-08-15 14:30 | Screenshot taken | Tester | SQL injection proof |
| 2026-08-15 14:31 | File hashed | Tester | SHA256: abc123… |
| 2026-08-15 14:35 | Stored securely | Tester | Encrypted volume |
8. Common Evidence Mistakes
Mistake 1: Insufficient Evidence
Bad: One screenshot showing the exploit command but not the result.
Good: Multiple screenshots showing the before state, the exploit, and the after state.
Mistake 2: Poor Quality Screenshots
Bad: Blurry, cropped, or unreadable screenshots.
Good: Clear, full-context screenshots with annotations.
Mistake 3: Missing Context
Bad: Screenshot of output without showing the command that produced it.
Good: Screenshot showing both the command and its output.
Mistake 4: No Flag Capture
Bad: Showing root access but not reading the flag file.
Good: Showing root access AND reading root.txt/proof.txt.
Mistake 5: Inconsistent Naming
Bad: screenshot1.png, screenshot2.png, img.png
Good: VULN-001_sqli-payload.png, VULN-001_sqli-result.png
9. OSCP Evidence Checklist
For Each Machine
□ Screenshot of initial reconnaissance (nmap output)
□ Screenshot of vulnerability identification
□ Screenshot of exploitation command
□ Screenshot of successful shell access
□ Screenshot of initial user context (whoami/id)
□ Screenshot of reading user.txt
□ Screenshot of privilege escalation method
□ Screenshot of escalated user context
□ Screenshot of reading root.txt/proof.txt
□ Screenshot of any lateral movement (if applicable)
10. Key Takeaways
- Evidence is mandatory. Every finding must be supported by clear, readable evidence.
- Take screenshots as you work. Don’t wait until the end to collect evidence.
- Show the full picture. Include before, during, and after states.
- Annotate your screenshots. Use arrows, boxes, and labels to highlight key information.
- Use consistent naming. Follow a clear naming convention for all evidence files.
- Show flag capture. Always include screenshots of reading user.txt and root.txt/proof.txt.
- Include command context. Show the command prompt and full output, not just partial results.
- For OSCP: Evidence is a graded component. Poor evidence can result in point deductions.
11. Further Reading
- NIST SP 800-72 Guide to Security Logging and Monitoring
- SANS Forensic Evidence Collection Guidelines
- ACPO (National Police Chiefs’ Council) Guidelines for Digital Evidence