Agent Guardrails MCP server

Read from the package source without running it. The installed server may list more.

• finish_agent_native_loop

  Finalize an agent-native loop by updating evidence, running check, and returning a reviewer summary.

• plan_rough_intent

  Parse a rough/fuzzy task description and generate a suggested task contract with inferred scope, test commands, and risk level. Use this when users provide vague or imprecise task descriptions.

• read_repo_guardrails

  Read the repo-local guardrail config, defaults, and policy summary.

• run_guardrail_checkwrite action

  Run the current guardrail check and return the structured result.

• start_agent_native_loop

  Bootstrap an agent-native task loop, write the runtime-backed contract, and seed evidence files.

• suggest_task_contract

  Bootstrap a task contract and runtime session from a natural-language task request.

• summarize_review_risks

  Summarize a prior guardrail check result into status, top risks, and next actions.

• Code scan32 source files scanned25/25
• –Live reliabilityno gateway calls yet and no remote to proben/a
• –Tool poisoningtools not inspected (local package is not executed); not countedn/a
• Auth qualitylocal package, no credentials required12/15
• Maintenancelast push 42 days ago12/15
• Maintainer identityregistry namespace matches repository owner; GitHub account older than a year8/10

Safety score

88/100

higher than 80% of 4,460 AI and agents servers

Tools

7

the typical AI and agents server has 6

Tools that change things

14%

typical: 0%. Fewer write tools means less an agent can break

GitHub stars

28

more than 90% of 2,711 AI and agents servers

Downloads a week

152

more than 71% of 1,538 AI and agents servers with a package

Agent Guardrails: common questions

Is Agent Guardrails MCP server safe?

Yes, by our scan: it is graded A (88/100). Read the Agent Guardrails safety report

How do I install Agent Guardrails?

It runs on your machine. Copy the Claude Code, Cursor, VS Code or Claude Desktop config from the install section.

Does Agent Guardrails need an API key?

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

Is Agent Guardrails maintained?

The last commit was 80 days ago (2026-08-16). The latest release is v0.3.3.

What can I use instead of Agent Guardrails?

Servers from other publishers that do the same job: Vibe Check MCP server, Serena MCP: the IDE for your agent MCP server and Figma Context MCP server.

How does Agent Guardrails compare with other AI and agents MCP servers?

safety score 88/100 (higher than 80% of 4,460 AI and agents servers); tools 7 (the typical AI and agents server has 6); tools that change things 14% (typical: 0%. Fewer write tools means less an agent can break). Full comparison.