LIGHTWEIGHT STATIC
CODE ANALYSIS WITH
SEMGREP
Nikhil Sahoo

WHOAMI
SECURITY SOFTWARE
ENGINEER
@MICROSOFT
EX- SECURITY
CONSULTANT,
DELOITTE
CERTS: OSCP,
EWPTX, CRTP
ACKNOWLEDGED
AND RECEIVED
HOFS FROM
MULTIPLE ORGS:
ORACLE, DELL,
MICROSOFT, SAP,
SONY ETC.
LINKEDIN:
HTTPS://WWW.LINKE
DIN.COM/IN/NIKHIL-
SAHOO-87204B106/
NIKHILSAHOO.COM

AGENDA
• History
• What is Semgrep?
• SAST Tools Slider
• Rule Syntax
• Writing Rules | Demo
• Usage
• Limitations
• QnA

HISTORY: SGREP (SYNTACTIC GREP)
• Initially called Sgrep/Pfff
• Written By Yoann Padioleau at Facebook for analyzing PHP code
• Was used to Enforce Best Practices
• Easy for developers to organize and understand the rules
• Joined R2C and renamed Sgrep to Semgrep
• Goal was to match based on semantics of the code
Reference: https://semgrep.dev/blog/2021/semgrep-a-static-analysis-journey

WHAT IS SEMGREP?
• Fast and lightweight static analysis tool to find bugs and enforce code
standards.
Reference: https://semgrep.dev/

PROS
• Opensource
• Rules look like the code you already write; no abstract syntax trees, regex
wrestling
• Scans in minutes
• Does not require a build
• Hunt Bugs at scale
• Supports many languages
• Very easy to write custom rules

SAST SLIDER
https://instagram-engineering.com/static-analysis-at-scale-an-instagram-story-8f498ab71a0c
Devskim
Semgrep
CodeQl

REGEX PROBLEMS
https://www.explainxkcd.com/wiki/images/1/10/perl_problems.png
False
Positives

REGEX PROBLEMS

CODE IS NOT A STRING, IT’S A TREE
🧶 String != 🌲Tree

INTERNALS
https://r2c.dev/static/00125f77fba64f5350b367c373c4e849/1132d/semgrep-flow.png

RULE SYNTAX
https://semgrep.dev/docs/writing-rules/rule-syntax/

DEMO WRITING SEMGREP RULES
HTTPS://SEMGREP.DEV/PLAYGROUND/NEW

ELLIPSIS
• Ellipsis Operator: “…”
• Find all PHP Command Injection Functions
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-php-
command-inj
• Reference: https://www.stackhawk.com/blog/php-command-injection/

METAVARIABLE
• Metavariable Operator: “$X”
• Find all user exploitable PHP Preg_Replace Functions
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-
metavariable-pregreplace
• Reference:
https://bitquark.co.uk/blog/2013/07/23/the_unexpected_dangers_of_preg
_replace

CATCHING ATTRIBUTES
• Scenario 1: Find all functions vulnerable to CSRF
• Solution: https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-
csrf-mvc
• Scenario 2: Broken Access Control
• Solution: https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-
missing-or-broken-authorization
• Reference: https://www.stackhawk.com/blog/net-csrf-protection-guide-
examples-and-how-to-enable/, https://github.com/returntocorp/semgrep-
rules/blob/develop/csharp/dotnet/security/audit/missing-or-broken-
authorization.cs

DEEP EXPRESSION
• Deep Expression Operator: “<… [your pattern]…>”
• Scenario: XML Parser allows resolving external resources
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-
xmltextreader
• Reference: https://rules.sonarsource.com/csharp/RSPEC-2755,
https://pvs-studio.com/en/blog/posts/csharp/0918/

METAVARIABLE REGEX
• Scenario: App not validating the expiration time of a JWT
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-jwt-no-
expiry
• Reference: https://github.com/returntocorp/semgrep-
rules/blob/develop/csharp/lang/security/ad/jwt-
tokenvalidationparameters-no-expiry-validation.cs

METAVARIABLE COMPARISON
• Scenario: Improper Regex Timeout
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-regex-
dos-infinite-timeout
• Reference: https://github.com/returntocorp/semgrep-
rules/blob/develop/csharp/lang/security/regular-expression-dos/regular-
expression-dos-infinite-timeout.cs,
https://rules.sonarsource.com/csharp/RSPEC-6444

TAINT ANALYSIS
• Scenario: PHP Reflected XSS
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-php-xss-
taint
• Reference: https://rules.sonarsource.com/php/RSPEC-5131

PATTERN-SANITIZERS
• Scenario: Path.Combine() Path Traversal
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-path-
combine
• Reference: https://www.praetorian.com/blog/pathcombine-security-
issues-in-aspnet-applications/

GENERIC PATTERN MATCHING
• Scenario: Debugging Enabled(Web.Config)
• Solution:
https://semgrep.dev/orgs/nikhil1232/editor/s/nikhil1232:owasp-net-
webconfig-debug
• Reference: https://portswigger.net/kb/issues/00100800_asp-net-
debugging-enabled

LIMITATION
Multi File Analysis:
DeepSemgrep

USAGE
Semgrep in Editors
• IntelliJ IDEA: semgrep-idea-plugin
• Microsoft Visual Studio Code: semgrep-vscode
• Vim: semgrep.vim
CLI
Docker

SEMGREP ON PUSH
https://semgrep.dev/docs/semgrep-ci/overview/

TOOLS
• Semgrep: https://github.com/returntocorp/semgrep
• Semgrep External Rules: https://github.com/returntocorp/semgrep-rules
• Semgrep Playground: https://semgrep.dev/playground
• CodeQL: https://codeql.github.com/
• Devskim: https://github.com/microsoft/DevSkim

REFERENCES
• https://semgrep.dev/docs/
• https://www.youtube.com/watch?v=kb8oo7Wyk84
• https://youtube.com/watch?v=O5mh8j7-An8
• https://semgrep.dev/playground/
• https://rules.sonarsource.com/
• https://semgrep.dev/blog/2021/semgrep-a-static-analysis-journey
• https://github.com/returntocorp/semgrep-rules

THANK YOU