Discover +86 AI Security apps & tools
Pros: Generates an AI Bill of Materials listing agents, tools, and credentials. Scans Terraform and CloudFormation templates for IaC misconfigurations. Provides a runtime gateway to monitor and control agent behavior. Self-hosted deployment via Docker keeps security data on your infrastructure.
Cons: Designed primarily for MCP environments, limiting non‑MCP applicability. Self-hosting requires internal operations and ongoing maintenance. CI/CD focus on GitHub Actions and Docker requires pipeline adaptation.
Pros: Standardized MCP interface for direct AI-agent connectivity. Integrates established tools such as Nmap, ffuf, and Nuclei. Extensible module architecture that accepts custom scripts. Runs on the host via Node.js, using local tooling.
Cons: Requires preinstallation of CLI security tools in the system PATH. Needs an MCP-compliant client and configuration to operate. Geared toward experienced security practitioners and researchers. Agent-driven actions and interpretations require human verification.
Pros: Local server keeps processed data inside the user's environment. MCP interface lets AI clients call operations during conversation. Includes standard hashing algorithms and AES encryption operations. Curated operation set exposes commonly used CyberChef functions to clients.
Cons: Requires a working Node.js installation and MCP-compatible client. Not every CyberChef operation from the full library is exposed. AI client's interpretation of results must be independently verified. Initial setup requires editing client configuration files.
Pros: Real-time security scanning for AI agent inputs and outputs. Detection of prompt injection and jailbreak attempts. PII detection and filtering to reduce data leakage risk. Open-source code and community-driven signature model.
Cons: Requires an MCP-compliant host and Node.js runtime. Optimized for agentic workflows, less relevant for simple LLM assistants. Deployment requires repository clone and manual MCP configuration.
Pros: MCP-native server integrates directly with clients like Claude Desktop. Renames identifiers to reduce human readability of Python source. Strips comments and docstrings to remove non-functional metadata. Preserves execution semantics so obfuscated scripts still run.
Cons: Python-only focus excludes non-Python projects. Requires an MCP-compatible host and local Python environment. Obfuscation is irreversible, complicating post-deployment debugging. Not a complete substitute for legal intellectual-property protections.
Pros: Detects and masks common PII types including emails and phone numbers. Processes input locally, avoiding cloud-side exposure to external AI providers. Configurable masking rules and open-source code allow security audits.
Cons: Requires MCP-compatible clients, limiting adoption to MCP-enabled workflows. Needs developer setup and a Node.js environment for deployment. Detection accuracy depends on rule configuration; human review advised.
Pros: Protocol-specific proxy designed for the Model Context Protocol. Inspects MCP request and response streams for tool-call visibility. Policy-based access control allows administrator-defined execution rules. Open-source GitHub repository enables community inspection and customization.
Cons: Requires Node.js deployment and environment familiarity. Policy definitions demand administrative setup and ongoing maintenance. Targeted at MCP adopters rather than general-purpose proxy users.
Pros: Enforces parameter-only model interaction, keeping raw credentials out of LLM inputs. Uses OS keychain for local secret storage and system-level secret injection. Native support for HTTP, GraphQL, and gRPC broadens backend compatibility. Acts as an MCP server for integration with MCP-compliant agents.
Cons: CLI and HCL template workflow requires technical engineering ownership. Local-first keychain model reduces centralized cloud secret store features. Requires template governance and logging to avoid misconfiguration.
Pros: Implements OAuth 2.1 for authorizing connections. Supports the Model Context Protocol (MCP) for AI-to-AI communication. Manages multiple OpenClaw instances with isolated authentication tokens. Docker-ready deployment for repeatable containerized environments.
Cons: Requires an MCP-compatible client such as Claude Desktop. Deployment outside Docker needs additional configuration effort. Setup and instance management require engineering resources.
Pros: Exposes security checks as standard MCP tools for native client calls. Detects embedded secrets and flags PII before model processing. Open-source architecture enables adding modules and integrations. Configurable security policies to tailor violation thresholds.
Cons: Malware scanning relies on third-party API keys such as VirusTotal. Requires hosting and maintaining a Python-based server. External scan accuracy depends on integrated service responses.
Pros: Executes Python and JavaScript/Node.js scripts for agent workflows. Configurable resource limits prevent runaway processes and excessive memory use. Open-source code base allows community auditing of sandbox mechanisms. Integrates with MCP clients via standard mcp_config.json configuration.
Cons: Requires a Node.js runtime and MCP-compatible client to run. Language support focused on scripting runtimes, primarily Python and JavaScript. Local server setup and configuration require developer knowledge.
Pros: Bridges FOFA search into AI workflows via the Model Context Protocol. Produces structured host metadata and basic statistical summaries. Open-source implementation recognized within the security researcher community.
Cons: Requires a FOFA account and API credentials as environment variables. Needs an MCP-compatible client and Node.js runtime. Search results depend on external index coverage and need verification.