Skip to main content
Glama
Mhdd-24

ai-guardrail-mcp

by Mhdd-24
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Hybrid

Related Servers

Alternatives to ai-guardrail-mcp

No user-submitted related servers found.

    Related Servers

    • sentinel-scan

      Ventrova
      A
      license
      Not graded
      quality
      B
      maintenance
      Enables MCP clients to scan tool manifests and server configs for prompt injection, tool poisoning, excessive agency, and other OWASP-mapped security issues directly through the scan_mcp_manifest tool.
      82 npm
      1
      MIT
    • MCP Security scanner by DeepSleuth

      MCP Security scannerofficial

      DeepSleuth
      A
      license
      A
      quality
      A
      maintenance
      Enables users to deterministically audit MCP servers for security issues across manifests, source, runtime responses, multi-call state, identity, and install scripts, and to run an inline gateway that gates tool calls and scans responses before returning them.
      3
      1
      Apache 2.0

    • toolfence

      dhb520cat
      A
      license
      A
      quality
      A
      maintenance
      Enables agents and CI pipelines to audit MCP servers and agent tool-chains by statically scanning repositories, local checkouts, tools/list exports, or live endpoints for risks such as destructive actions without confirmation, mismatched safety annotations, injection surfaces, credential or PII exposure, and unguarded command, path, or URL sinks. All checks are read-only and never execute the scanned code or call tools/call.
      3
      MIT
    Deploy Server
    One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

    TDQS

    B3.2/5.0

    Scored across 3 tools

    Disambiguation5/5

    The three tools have clearly distinct purposes: status/health checking, text scanning, and test suite generation. There is no overlap or ambiguity between them, making tool selection straightforward.

    Naming Consistency5/5

    All tool names follow a uniform aiguard_ prefix pattern with a single lowercase descriptive word. This consistent naming convention makes the tool set predictable and easy to navigate.

    Tool Count4/5

    Three tools is a compact but reasonable size for a focused guardrail utility. Each tool has a distinct purpose, though the set could potentially be slightly expanded with additional related operations without feeling bloated.

    Completeness3/5

    The tools cover health/status, text scanning, and test outline generation, but notable gaps exist—there is no policy management or execution capability for the generated test suite. Specifically, aiguard_suite creates an outline with no way to run it, creating a dead end in the workflow.

    Maintenance

    ActivityMaintained
    ResponsivenessNo issues

    Resources