### Example Apt Get Install Command

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows an apt-get install command that mantishack-sca recognizes when parsing inline install sources.

```bash
apt-get install -y curl
```

--------------------------------

### Example Dockerfile Inline Install

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows how mantishack-sca parses inline installation commands within Dockerfiles.

```dockerfile
FROM ubuntu:latest
RUN apt-get update && apt-get install -y curl
RUN curl -fsSL https://example.com/install.sh | bash
```

--------------------------------

### Scenario: General Help

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Example command to get help securing an application using MANTISHACK.

```plaintext
/mantishack - help me secure my application
```

--------------------------------

### Example Gem Install Command

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows a gem install command that mantishack-sca recognizes when parsing inline install sources.

```bash
gem install rails
```

--------------------------------

### Example Fuzzing Workflow

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/FUZZING_QUICKSTART.md

Demonstrates a typical workflow for fuzzing a binary with MantiHack, starting with a quick smoke test and progressing to deeper goal-directed analysis.

```bash
python3 mantishack_fuzzing.py \
    --binary ./target_app \
    --duration 300 \
    --autonomous \
    --max-crashes 5
```

```bash
python3 mantishack_fuzzing.py \
    --binary ./target_app \
    --duration 3600 \
    --autonomous \
    --goal "find heap overflow" \
    --parallel 4 \
    --max-crashes 20
```

```bash
ls out/fuzz_*/autonomous_corpus/    # View generated seeds
cat out/fuzz_*/fuzzing_report.json  # Summary report
ls out/fuzz_*/analysis/exploits/     # Generated exploits
```

--------------------------------

### Run SAGE Setup Script

Source: https://github.com/deonmenezes/mantishack/blob/main/core/sage/docs/SAGE_INTEGRATION.md

Execute the setup script to configure SAGE integration. This script handles SDK verification, MCP configuration, Docker setup, knowledge seeding, and agent registration.

```bash
libexec/mantishack-sage-setup
```

--------------------------------

### Example Npm Install Command

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows an npm install command that mantishack-sca recognizes when parsing inline install sources.

```bash
npm install lodash
```

--------------------------------

### Example Python Pip Install Command

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows a pip install command that mantishack-sca recognizes when parsing inline install sources.

```bash
pip install requests
```

--------------------------------

### Example Cargo Install Command

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows a cargo install command that mantishack-sca recognizes when parsing inline install sources.

```bash
cargo install ripgrep
```

--------------------------------

### Example .NET Project File

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet is an example of a .csproj file for a .NET project, which mantishack-sca scans for package references.

```xml
<Project Sdk="Microsoft.NET.Sdk">
  <PropertyGroup>
    <OutputType>Exe</OutputType>
    <TargetFramework>net6.0</TargetFramework>
  </PropertyGroup>
  <ItemGroup>
    <PackageReference Include="Newtonsoft.Json" Version="13.0.1" />
  </ItemGroup>
</Project>
```

--------------------------------

### Install Mantishack Manually

Source: https://github.com/deonmenezes/mantishack/blob/main/README.md

Follow these steps to clone the repository, install Python dependencies, and set up required tools like Claude Code and Semgrep for manual installation.

```bash
git clone https://github.com/deonmenezes/mantishack.git
cd mantishack

pip install -r requirements.txt

npm install -g @anthropic-ai/claude-code

pip install semgrep

claude
```

--------------------------------

### Install SAGE Agent SDK and httpx

Source: https://github.com/deonmenezes/mantishack/blob/main/requirements.txt

Install the SAGE agent SDK and httpx for SAGE persistent memory functionality.

```bash
pip install sage-agent-sdk==6.6.5 httpx==0.28.1
```

--------------------------------

### Example Security Scan Workflow

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_USAGE.md

Demonstrates an interactive workflow starting with a /scan command, followed by analysis and user choices for remediation.

```bash
User: /scan

Claude: [Python executes mantishack.py scan]
        [Scan completes]

        Results: 15 vulnerabilities found

        Analyzing with adversarial thinking:
        - 2 Secrets (instant compromise) - REVIEW FIRST
        - 5 SQLi (database access) - REVIEW SECOND
        - 3 Auth issues (broken access control) - REVIEW THIRD
        - 5 Config issues - REVIEW LAST

        What next?
        1. Deep - Analyze top findings
        2. Fix - Apply patches/improve exploits
        3. Generate report - Export results
        4. Retry - Different parameters
        5. Done - Finish

        Your choice? [1-5]

User: 2

Claude: Reviewing Python-generated patches...
        Found 2 template patches. Using Patch Engineer persona to create actual fixes...
        ✓ All patches now production-ready
        Apply patches? [Y/n]

User: Y

Claude: [Applies patches using Edit tool]
        ✓ Patches applied successfully
        Run /scan again to verify fixes? [Y/n]
```

--------------------------------

### MANTISHACK Setup in Claude Code

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Steps to set up MANTISHACK within Claude Code, including cloning the repo, installing dependencies, and configuring LLM access.

```bash
# 1. Clone and open in Claude Code
git clone https://github.com/gadievron/raptor.git
cd mantishack
claude

# 2. Let Claude handle setup
"Install Python packages from requirements.txt"
"Install semgrep"  # External tool

# 3. Set up LLM (choose one)
"Set my ANTHROPIC_API_KEY to [your-key]"          # Cloud (best quality)
# OR
"Install Ollama and pull deepseek-r1 model"       # Local/free

# 4. Start using MANTISHACK
/scan - Scan code for vulnerabilities
/fuzz - Fuzz binaries (asks to install AFL++ if needed)
/web  - Test web applications
```

--------------------------------

### Example Dotnet Add Package Command

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows a dotnet add package command that mantishack-sca recognizes when parsing inline install sources.

```bash
dotnet add package Newtonsoft.Json
```

--------------------------------

### Install SAGE SDK

Source: https://github.com/deonmenezes/mantishack/blob/main/core/sage/docs/SAGE_INTEGRATION.md

Install the necessary Python packages for SAGE integration. This includes the SDK and httpx for HTTP requests.

```bash
pip install sage-agent-sdk httpx
```

--------------------------------

### Example Go Module File

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet is an example of a go.mod file for a Go project, which mantishack-sca scans for module dependencies.

```go
module example.com/my-go-app

go 1.18

require (
	r) github.com/gin-gonic/gin v1.8.1
)
```

--------------------------------

### Example Dataflow Analysis Prompt

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/PHASE2_OVERVIEW.md

This prompt guides an LLM to analyze a CodeQL dataflow path, assessing sanitizer effectiveness, bypass possibilities, and attack complexity.

```plaintext
You are analyzing a dataflow path:

SOURCE: User input from HTTP parameter
  LoginController.java:45

STEP 1: Passes through validation
  InputValidator.java:23

SINK: SQL query execution
  UserDAO.java:78

SANITIZERS: Basic input validation

Determine:
1. Can the sanitizer be bypassed?
2. Is this truly exploitable?
3. What's the attack complexity?
```

--------------------------------

### Install MANTISHACK Framework

Source: https://github.com/deonmenezes/mantishack/blob/main/requirements.txt

Install the MANTISHACK framework and its core dependencies using pip.

```bash
pip3 install -r requirements.txt
```

--------------------------------

### Install Optional LLM Provider SDKs

Source: https://github.com/deonmenezes/mantishack/blob/main/requirements.txt

Install SDKs for various LLM providers to enable external LLM support. Choose the SDK that matches your provider.

```bash
pip install openai        # OpenAI, Gemini (via shim), Mistral, Ollama
```

```bash
pip install anthropic     # Anthropic Claude (native structured output)
```

```bash
pip install google-genai  # Google Gemini (native SDK, accurate thinking token costs)
```

--------------------------------

### Install Graphviz on Ubuntu/Debian

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/DATAFLOW_VISUALIZATION.md

Install the Graphviz package using apt-get on Ubuntu or Debian systems.

```bash
# Ubuntu/Debian
sudo apt-get install graphviz
```

--------------------------------

### Install Optional Dataflow Visualization Tool

Source: https://github.com/deonmenezes/mantishack/blob/main/requirements.txt

Install the tabulate package for enhanced dataflow visualization, which is recommended for better analysis.

```bash
pip install tabulate==0.10.0
```

--------------------------------

### Audit Budget CLI Examples

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sandbox.md

Examples of how to configure the audit budget via the command-line interface for different diagnostic needs. Adjust the budget value based on workload requirements.

```bash
mantishack scan target/  --sandbox full --audit                    # default 10000
mantishack scan target/  --sandbox full --audit --audit-budget 100 # quick diag
mantishack scan target/  --sandbox full --audit --audit-verbose --audit-budget 50000  # long run
```

--------------------------------

### Install GDB

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/FUZZING_QUICKSTART.md

Installs the GNU Debugger (GDB). Use the command suitable for your OS (macOS or Ubuntu).

```bash
# macOS
brew install gdb

# Ubuntu
sudo apt install gdb
```

--------------------------------

### Install Optional Rich Inventory Metadata Tools

Source: https://github.com/deonmenezes/mantishack/blob/main/requirements.txt

Install packages for enhanced metadata handling, including decorators, annotations, and typed parameters.

```bash
pip install tree-sitter tree-sitter-python tree-sitter-java tree-sitter-javascript tree-sitter-c tree-sitter-go
```

--------------------------------

### Install Z3 SMT Solver

Source: https://github.com/deonmenezes/mantishack/blob/main/README.md

Command to install the Z3 SMT solver library, which is optional but enhances Mantishack's analysis results.

```bash
pip install z3-solver
```

--------------------------------

### Example Java Maven POM File

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet is an example of a pom.xml file for a Java project using Maven, which mantishack-sca scans.

```xml
<project>
  <modelVersion>4.0.0</modelVersion>
  <groupId>com.example</groupId>
  <artifactId>my-app</artifactId>
  <version>1.0-SNAPSHOT</version>
  <dependencies>
    <dependency>
      <groupId>junit</groupId>
      <artifactId>junit</artifactId>
      <version>4.13.2</version>
      <scope>test</scope>
    </dependency>
  </dependencies>
</project>
```

--------------------------------

### Verify Mantishack Installation and Components

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/ARCHITECTURE.md

Commands to verify the installation of the main Mantishack launcher and its individual components, including static analysis, CodeQL, LLM analysis, and full workflows.

```bash
# Test main launcher
python3 mantishack.py

# Test static analysis
python3 packages/static-analysis/scanner.py --help

# Test CodeQL
python3 mantishack_codeql.py --help

# Test LLM analysis
python3 packages/llm_analysis/agent.py --help

# Test full workflows
python3 mantishack_agentic.py --help
python3 mantishack_fuzzing.py --help
```

--------------------------------

### Patch Output Diff Example

Source: https://github.com/deonmenezes/mantishack/blob/main/tiers/personas/patch_engineer.md

An example of the expected patch output format, including a diff, comments explaining the fix, and vulnerability type reference. Include clear diff format, comments explaining the fix, reference to vulnerability type, and testing recommendations.

```diff
--- vulnerable_file.py (original)
+++ vulnerable_file.py (patched)
@@ -45,3 +45,4 @@
def login(username, password):
-    query = f"SELECT * FROM users WHERE name='{username}'"
+    # Fixed: Use parameterized query to prevent SQL injection
+    query = "SELECT * FROM users WHERE name=?"
+    cursor.execute(query, (username,))
```

--------------------------------

### Usage Examples for Dependency Scan

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/EXTENDING_LAUNCHER.md

Demonstrates how to execute the dependency scan mode from the command line and access help information.

```bash
# Run dependency scan
python3 mantishack.py depscan --repo /path/to/code

# Get help
python3 mantishack.py help depscan

```

--------------------------------

### MantiHack CodeQL CLI Usage Example

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/PHASE2_OVERVIEW.md

Demonstrates how to run MantiHack CodeQL in fully autonomous mode using the command-line interface, specifying the repository path.

```bash
# Fully autonomous (zero configuration)
python3 mantishack_codeql.py --repo /path/to/code
```

--------------------------------

### Scenario: Find Binary Bugs

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Example command to fuzz a binary for a specified duration using MANTISHACK.

```plaintext
/mantishack-fuzz - fuzz ./myapp for 30 minutes
```

--------------------------------

### Static Analysis Scanner CLI Example

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/ARCHITECTURE.md

Demonstrates how to run the static analysis scanner package from the command line, specifying the repository path, policy groups, and output directory.

```bash
python3 packages/static-analysis/scanner.py \
  --repo /path/to/code \
  --policy_groups secrets,owasp \
  --output /path/to/output
```

--------------------------------

### Patch Example: MD5 vs. SHA-256

Source: https://github.com/deonmenezes/mantishack/blob/main/tiers/personas/exploit_developer.md

Demonstrates how to correctly apply a patch by showing the actual code change for replacing MD5 with SHA-256.

```diff
- digest = MessageDigest.getInstance("MD5");
+ digest = MessageDigest.getInstance("SHA-256");
```

--------------------------------

### Scanner.py Help Text Example

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/ARCHITECTURE.md

Demonstrates how to use the --help flag for the static analysis scanner to understand its arguments and usage.

```bash
$ python3 packages/static-analysis/scanner.py --help

MANTISHACK Static Analysis Scanner

Scans code using Semgrep with configurable policy groups.

Required Arguments:
  --repo PATH          Path to repository to scan

Optional Arguments:
  --policy_groups STR  Comma-separated policy groups (default: all)
  --output PATH        Output directory (default: auto-generated)

Examples:
  # Scan with all policy groups
  python3 scanner.py --repo /path/to/code

  # Scan specific policy groups
  python3 scanner.py --repo /path/to/code --policy_groups secrets,owasp
```

--------------------------------

### Example Composer Require Command

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet shows a composer require command that mantishack-sca recognizes when parsing inline install sources.

```bash
composer require monolog/monolog
```

--------------------------------

### CodeQL Pack Structure Example

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/llm_analysis/codeql_packs/README.md

Illustrates the directory layout for a CodeQL pack, including the qlpack.yml, codeql-pack.lock.yml, and the Mantishack/LocalFlowSource.qll library.

```yaml
<lang>-queries/
├── qlpack.yml              — depends on codeql/<lang>-all: "*"
├── codeql-pack.lock.yml    — committed; deps pinned for reproducibility
├── Mantishack/
│   └── LocalFlowSource.qll — selects threat-model sources
└── Security/
    ├── CWE-022/PathTraversalLocal.ql
    ├── CWE-078/CommandInjectionLocal.ql
    └── ...
```

--------------------------------

### Run CodeQL Analysis with Caching

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/README.md

Demonstrates the difference in execution time between the first run and a subsequent run when caching is utilized. The first run creates the database, while the second run leverages the cache for faster analysis.

```bash
# First run
python3 packages/codeql/agent.py --repo /path/to/code
# Database created: 8.3 minutes

# Second run (cache hit)
python3 packages/codeql/agent.py --repo /path/to/code
# ✓ Using cached database for java (0.2 seconds)
# Total time: 2.1 minutes (just query execution)
```

--------------------------------

### Install MANTISHACK Dependencies

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/README.md

Installs the necessary Python dependencies and static analysis tools for MANTISHACK. Ensure you have Python 3.10+ and Git installed.

```bash
# Clone repository
git clone <repo-url>
cd MANTISHACK-daniel-modular

# Install Python dependencies
pip3 install anthropic openai requests beautifulsoup4 pwntools tabulate

# Install static analysis tools
pip3 install semgrep

# Download and configure CodeQL
# See: https://codeql.github.com/docs/codeql-cli/getting-started-with-the-codeql-cli/

# Install AFL++ (for binary fuzzing)
brew install afl++  # macOS
# OR
sudo apt install afl++  # Ubuntu

# Install GDB (for crash analysis)
brew install gdb  # macOS
# OR
sudo apt install gdb  # Ubuntu

# Verify installation
python3 mantishack_agentic.py --help
python3 mantishack_fuzzing.py --help
```

--------------------------------

### Analyze Java Maven Project

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/README.md

Example of running CodeQL analysis on a Java project using Maven. Shows expected output including detected language, build system, database creation time, and findings.

```bash
python3 packages/codeql/agent.py \
  --repo /Users/daniel/O365/CSR/Code/CodeQL-Crypto-Research/GHUniverse/acme-access-main

# Output:
# ✓ Detected java (confidence: 0.92)
# ✓ Detected Maven build system
# ✓ Created database (5.2 minutes)
# ✓ Analysis completed: 12 findings
# SARIF: out/codeql_acme-access-main_20250114_123456/codeql_java.sarif
```

--------------------------------

### Install Python Dependencies for Mantishack

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/ARCHITECTURE.md

Installs all Python dependencies required for the Mantishack project using pip and a requirements file. Alternatively, individual packages can be installed.

```bash
git clone <repo-url>
cd mantishack

pip install -r requirements.txt

# Or install manually:
pip install semgrep anthropic openai instructor requests beautifulsoup4
```

--------------------------------

### Analyze Multi-Language Repository

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/README.md

Example of running CodeQL analysis on a repository containing multiple programming languages. Demonstrates parallel database creation and aggregated findings.

```bash
python3 packages/codeql/agent.py --repo /path/to/fullstack-app

# Output:
# ✓ Detected python (confidence: 0.85)
# ✓ Detected javascript (confidence: 0.78)
# ✓ Detected java (confidence: 0.91)
# ✓ Created 3 databases in parallel
# ✓ Analysis completed: 47 findings total
```

--------------------------------

### Install AFL++

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/FUZZING_QUICKSTART.md

Commands to install AFL++ on macOS and Ubuntu systems.

```bash
# Install AFL++
brew install afl++  # macOS
sudo apt install afl++  # Ubuntu
```

--------------------------------

### Structured Exploitation Paths Example

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/exploitability-validation-integration.md

Demonstrates the JSON structure for exploitation paths, pairing techniques with binary targets.

```json
{
  "exploitation_paths": [
    {"technique": "heap metadata corruption via double-free", "target": "GOT[exit] @ 0x28570"},
    {"technique": "GOT overwrite via heap metadata corruption", "target": "GOT[_ex_deregister] @ 0x28aa0"}
  ]
}
```

--------------------------------

### Example Ruby Gemfile

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet is an example of a Gemfile for a Ruby project, which mantishack-sca parses for gem dependencies.

```ruby
source 'https://rubygems.org'

gem 'rails', '~> 7.0.0'
gem 'pg', '~> 1.0'
```

--------------------------------

### Binary Fuzzing Natural Language Examples

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Examples of natural language prompts for fuzzing binaries with MANTISHACK.

```plaintext
/fuzz - fuzz this binary
/fuzz - test ./myapp for 5 minutes
/fuzz - find crashes in /usr/local/bin/myapp
```

--------------------------------

### Mantishack Project Setup and Scan

Source: https://github.com/deonmenezes/mantishack/blob/main/README.md

Commands to create a project, set it as active, map the attack surface, and initiate an agentic scan. The agentic scan includes Semgrep and CodeQL, audit lanes, and validation.

```bash
/mantis-project create myapp --target /path/to/code   # create a project first
/mantis-project use myapp                             # set it as active
/mantis-understand --map                              # map the attack surface
/mantis-agentic                                       # scan, audit, validate, exploit, patch
/mantis-project findings                              # review everything in one place
```

--------------------------------

### Install Graphviz on macOS

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/DATAFLOW_VISUALIZATION.md

Install the Graphviz package using Homebrew on macOS to enable DOT file rendering.

```bash
# macOS
brew install graphviz
```

--------------------------------

### Build CodeQL Databases for Multiple Languages

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/llm_analysis/tests/fixtures/iris_e2e_multilang/README.md

Commands to create CodeQL databases for JavaScript, Java, Go, and C fixtures. For Java and C, specific compilation commands are included.

```bash
codeql database create /tmp/iris-js   --language=javascript --source-root=js   --overwrite
```

```bash
codeql database create /tmp/iris-java --language=java       --source-root=java --command="javac CmdInj.java" --overwrite
```

```bash
codeql database create /tmp/iris-go   --language=go         --source-root=go   --overwrite
```

```bash
codeql database create /tmp/iris-c    --language=cpp        --source-root=c    --command="gcc -c cmd_inj.c -o /tmp/cmd_inj.o" --overwrite
```

--------------------------------

### Install AFL++

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/FUZZING_QUICKSTART.md

Installs AFL++ fuzzing tool. Use the appropriate command for your operating system (macOS or Ubuntu/Debian).

```bash
# macOS
brew install afl++

# Ubuntu/Debian
sudo apt install afl++

# Verify installation
which afl-fuzz
```

--------------------------------

### Example Rust Cargo.toml File

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet is an example of a Cargo.toml file for a Rust project, which mantishack-sca parses for dependencies.

```toml
[package]
name = "my-rust-app"
version = "0.1.0"
edition = "2021"

[dependencies]
serde = { version = "1.0", features = ["derive"] }
```

--------------------------------

### Expert Persona Usage Examples

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_USAGE.md

Examples of how to invoke expert personas for specific tasks. These personas are loaded on-demand.

```bash
"Use exploit developer persona to create PoC for finding #42"
"Use crash analyst persona to analyze this crash"
"Use fuzzing strategist persona to recommend AFL parameters"
```

--------------------------------

### Command-Based Input Detection Example

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/FUZZING_QUICKSTART.md

Illustrates binary analysis output when detecting commands and wrapping seeds, showing generated counts.

```text
[INFO] Detected command: STACK
[INFO] Detected command: HEAP
[INFO] Detected command: UAF
[INFO] Detected command: JSON
[INFO] Detected command: XML
[INFO] Binary analysis complete: 5 formats, 8 commands detected
[INFO] Wrapping basic seeds with 8 detected commands
[INFO] Generated 96 basic seeds  (12 × 8 commands)
[INFO] Wrapping goal-directed seeds with STACK command
[INFO] Generated 5 goal-directed seeds
```

--------------------------------

### Web Testing Natural Language Examples

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Examples of natural language prompts for testing web applications with MANTISHACK.

```plaintext
/web - test my website
/web - scan https://localhost:3000 for XSS
/web - check this app for OWASP Top 10
```

--------------------------------

### Create CodeQL Database for OWASP Benchmark Java

Source: https://github.com/deonmenezes/mantishack/blob/main/core/dataflow/corpus/SOURCES.md

Creates a CodeQL database for the OWASP Benchmark Java project, used for security analysis.

```bash
# 1. Clone (see above)
# 2. Build CodeQL DB (the build hits Maven, takes ~3-5 minutes)
codeql database create /tmp/owasp-codeql-db \
    --language=java \
    --command="mvn -B -DskipTests clean package" \
    --source-root=out/dataflow-corpus-fixtures/owasp-benchmark-java \
    --overwrite
```

--------------------------------

### Run CodeQL Analysis with Agent

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/ARCHITECTURE.md

Execute the CodeQL analysis agent with specified repository, language, and output path. Ensure the CodeQL CLI is installed.

```bash
python3 packages/codeql/agent.py \
  --repo /path/to/code \
  --language python \
  --output /path/to/output
```

--------------------------------

### Code Scanning Natural Language Examples

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Examples of natural language prompts for scanning source code with MANTISHACK.

```plaintext
/scan - scan this directory
/scan - find secrets in my code
/scan - check for SQL injection vulnerabilities
```

--------------------------------

### Install Python Dependencies

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/FUZZING_QUICKSTART.md

Installs necessary Python libraries for MANTISHACK, including libraries for network requests and LLM providers.

```bash
pip3 install requests anthropic openai pwntools
```

--------------------------------

### Compile and Analyze CodeQL Queries

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/llm_analysis/codeql_packs/README.md

Use these commands to compile a new CodeQL query and then analyze a database with it. Ensure the query compiles successfully and produces results on a known-vulnerable fixture.

```bash
codeql query compile path/to/NewLocal.ql
codeql database analyze <fixture-db> path/to/NewLocal.ql \
    --format=sarif-latest --output=/tmp/result.sarif
```

--------------------------------

### Profile System (Context-Aware Analysis)

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/exploit_feasibility/README.md

Demonstrates how to create and use profiles for context-aware analysis, especially for remote targets.

```APIDOC
## Profile System (Context-Aware Analysis)

### `create_remote_profile(glibc_version)`

#### Description
Creates a remote profile, optionally specifying the glibc version for accurate analysis.

#### Parameters
- **glibc_version** (string) - Optional - The version of glibc to use for the profile (e.g., "2.31").

### `FeasibilityAnalyzer(profile)`

#### Description
Initializes a FeasibilityAnalyzer with a given profile.

#### Parameters
- **profile** (object) - Required - The profile object to use for analysis.

### `FeasibilityAnalyzer.full_analysis()`

#### Description
Performs a full feasibility analysis using the configured analyzer and profile.

#### Returns
- **report** (object) - The analysis report.
```

--------------------------------

### Example PHP Composer JSON

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet is an example of a composer.json file for a PHP project, which mantishack-sca parses for package dependencies.

```json
{
    "require": {
        "monolog/monolog": "^2.0"
    }
}
```

--------------------------------

### Scenario: Check New Repo

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Example command to scan a newly cloned repository for security issues using MANTISHACK.

```plaintext
/mantishack-scan - scan this repository for all security issues
```

--------------------------------

### Scenario: Test Web App

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/CLAUDE_CODE_QUICKSTART.md

Example command to test a local web application before deployment using MANTISHACK.

```plaintext
/mantishack-web - test http://localhost:8000
```

--------------------------------

### Example Node.js Package Lock File

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sca.md

This snippet is an example of a package-lock.json file used by Node.js projects, which mantishack-sca parses.

```json
{
  "name": "my-app",
  "version": "1.0.0",
  "dependencies": {
    "lodash": {
      "version": "4.17.21",
      "resolved": "https://registry.npmjs.org/lodash/-/lodash-4.17.21.tgz"
    }
  }
}
```

--------------------------------

### Running MantiHack CodeQL Autonomous Workflow

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/PHASE2_OVERVIEW.md

Demonstrates how to set up an API key and execute the MantiHack CodeQL fully autonomous workflow for a given repository. Review generated exploits after execution.

```bash
# Set your API key
export ANTHROPIC_API_KEY=sk-...

# Run fully autonomous workflow
python3 mantishack_codeql.py --repo /path/to/your/java/project

# Wait 20-60 minutes
# Review exploits in out/codeql_*/autonomous/exploits/
```

--------------------------------

### Sandbox Context with Egress Proxy and Multiple Runs

Source: https://github.com/deonmenezes/mantishack/blob/main/docs/sandbox.md

This example demonstrates using a `sandbox` context manager with an egress proxy enabled. It shows how to make multiple calls within the same sandbox and access the combined event stream via `run.events`.

```python
with sandbox(use_egress_proxy=True, proxy_hosts=["api.example.com"]) as run:
    run(["curl", "https://api.example.com/a"])
    run(["curl", "https://api.example.com/b"])
    print(run.events)  # combined list covering both calls
```

--------------------------------

### Example Analysis JSON with Visualization Paths

Source: https://github.com/deonmenezes/mantishack/blob/main/packages/codeql/DATAFLOW_VISUALIZATION.md

This JSON structure shows an example of an analysis result that includes paths for various visualization formats.

```json
{
  "finding": {
    "rule_id": "java/sql-injection",
    "has_dataflow": true,
    "dataflow_path_count": 1,
    ...
  },
  "analysis": {
    "is_exploitable": true,
    ...
  },
  "dataflow_validation": {
    "is_exploitable": true,
    "confidence": 0.95,
    "sanitizers_effective": false,
    ...
  },
  "visualizations": {
    "html": "out/.../autonomous/visualizations/java_sql-injection_42_dataflow.html",
    "mermaid": "out/.../autonomous/visualizations/java_sql-injection_42_dataflow.mmd",
    "ascii": "out/.../autonomous/visualizations/java_sql-injection_42_dataflow.txt",
    "dot": "out/.../autonomous/visualizations/java_sql-injection_42_dataflow.dot"
  }
}
```