AI Guardrails MCP server

• check_hallucinationFree

  Heuristic scoring of LLM hedge phrases that correlate with hallucinated outputs. Returns score 0–1 and risk level.

• check_safetyFree

  Check text for harmful content using weighted keyword detection. Returns a 0–1 score and verdict (safe/caution/unsafe).

• detect_piiFree

  Detect personally identifiable information (email, phone, SSN, credit card, IP, URL, API key) in a text string.

• full_checkwrite actionFree

  Run pii, safety, and/or hallucination checks in one call. Default: pii + safety. Specify checks=["pii","safety","hallucination"] to run all.

• redact_piiFree

  Redact PII from text. mode=mask replaces with [REDACTED]; mode=tag uses typed placeholders like [EMAIL].

• –Code scanremote-only server, no package to scann/a
• Live reliabilityremote reachable in 129ms20/20
• Tool poisoning5 tool descriptions checked15/15
• Auth qualityopen endpoint exposes 1 write-action tools with no auth3/15
• Maintenancelast push 170 days ago8/15
• Maintainer identityregistry namespace matches repository owner6/10

Safety score

69/100

higher than 28% of 1,119 Developer tools servers

Tools

5

the typical Developer tools server has 7

Tools that change things

20%

typical: 5%. Fewer write tools means less an agent can break

Uptime

100%

of our recent checks; the typical remote Developer tools server answers 100%

Response time

21 ms

faster than 96% of 443 Developer tools servers

AI Guardrails: common questions

Is AI Guardrails MCP server safe?

With care: it is graded C, so read the findings first (69/100). Read the AI Guardrails safety report

How do I install AI Guardrails?

It runs remotely at api.lazy-mac.com. Add it to Claude Code, Cursor, VS Code or Claude Desktop with the snippets above, or call it through the mcp.market gateway without installing anything.

Does AI Guardrails need an API key?

Not as far as the registry entry and our scan can tell: no credentials are declared or required.

Is AI Guardrails maintained?

The last commit was 207 days ago (2026-04-11). The latest release is v1.0.0.

Is AI Guardrails up?

100% of our last 28 checks got an answer. We check remote servers about four times a day.

What can I use instead of AI Guardrails?

Servers from other publishers that do the same job: Scrapling MCP Server, Porkbun (official) MCP server and MCP server.

How does AI Guardrails compare with other Developer tools MCP servers?

safety score 69/100 (higher than 28% of 1,119 Developer tools servers); tools 5 (the typical Developer tools server has 7); tools that change things 20% (typical: 5%. Fewer write tools means less an agent can break). Full comparison.