MCP For Claude Marketplace
Template Injection Sanitizer

Template Injection Sanitizer MCP Connector for Claude

A+

Detects and neutralizes prompt injection attacks using template syntax like Jinja2 and Python f-strings.

3 tools Official Updated Oct 1, 2026 Official Vinkius Partner
More Details Connect to Claude

Protect your LLM applications from prompt injection attacks targeting template engines. This MCP server provides specialized tools to identify and remove malicious syntax used in frameworks like LangChain and CrewAI. Use scan_input_security to inspect individual strings for Jinja2 expressions, Python f-string patterns, or dangerous system variable access like __class__. For high-volume workloads, batch_sanitize_inputs allows for efficient bulk cleaning of user-generated content. This utility ensures that user inputs cannot manipulate prompt structures or leak sensitive data through interpolation.

securityprompt-injectionsanitizationjinja2pythonllm-safety

3 tools expose this connector's capabilities to your AI agent.

batch_sanitize_inputs

Processes a collection of strings to allow for bulk cleaning of user-generated content

get_injection_risk_profile

Evaluates the severity of an injection attempt based on the density and type of patterns found

scan_input_security

Analyzes a single input string to identify potential injection attempts and provides a cleaned version

See how to talk to your AI agent using Template Injection Sanitizer.

Check if this input is safe: 'Hello {{ user_name }}'

The input contains a Jinja2 expression and has been flagged as injected.

Is this string malicious? 'print(obj.__class__)'

The input contains system variable access and is considered a high-risk injection attempt.

Sanitize this list: ['safe text', '{secret_var}']

The first item is safe, and the second item was sanitized by removing the Python f-string pattern.

It prevents template-based prompt injection where attackers use Jinja2 or Python f-string syntax to manipulate LLM prompts or access system variables.

Related Connectors

Prompt Template Variable Resolver

Prompt Template Variable Resolver MCP

3 tools Official

Validates and resolves variable placeholders in LLM templates.

A+ View details →
Output Format Stability Checker

Output Format Stability Checker MCP

3 tools Official

Detect non-deterministic shifts in LLM response structures.

A+ View details →
LLM JSON Repair Utility

LLM JSON Repair Utility MCP

3 tools Official

Fixes malformed, truncated, or dirty JSON strings produced by LLMs.

A+ View details →
Password Manager Export Analyzer

Password Manager Export Analyzer MCP

1 tools Official

Analyze Bitwarden, LastPass, or 1Password CSV exports for weak and duplicate passwords — without EVER sending real passwords to the AI.

A+ View details →