Scan text (tool descriptions, tool responses, MCP server manifest content) for prompt injection attacks, hidden instructions, zero-width characters, jailbreak patterns, and instruction overrides.
AI agents call check_prompt_injection to retrieve information from Mcp Shield without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool performs static analysis and pattern matching on text content. It reads input (descriptions, responses, manifests) and reports findings but does not create, modify, delete, or execute anything. The blast radius of misuse is minimal—worst case it produces false positives or false negatives in security reports, neither of which changes system state.
From the tool's definition Tool scans text for security patterns without modifying data: 'Scan text (tool descriptions, tool responses, MCP server manifest content) for prompt injection attacks, hidden instructions, zero-width characters, jailbreak patterns, and instruction overrides.'…
Attacks that exploit this kind of access
Scan text (tool descriptions, tool responses, MCP server manifest content) for prompt injection attacks, hidden instructions, zero-width characters, jailbreak patterns, and instruction overrides. It is categorised as a Read tool in the Mcp Shield MCP Server, which means it retrieves data without modifying state.
Register the Mcp Shield MCP server in PolicyLayer and add a rule for check_prompt_injection: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Mcp Shield. Nothing to install.
check_prompt_injection is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the check_prompt_injection rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.
Set action: deny in the PolicyLayer policy for check_prompt_injection. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.
check_prompt_injection is provided by the Mcp Shield MCP server (muhannad-hash/mcp-shield). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →