Arcezia / Integrations

Integrations: check agent tool calls in the framework you use

Each hookup checks an AI agent’s tool call before it runs and lets only ALLOW reach the tool. Install with pip install arcezia (Python 3.10 or newer); any language can call the HTTP API.

• Claude Code: a PreToolUse hook that checks shell commands, writes, web access and MCP tools.
• LangChain and LangGraph: wrapped tools that raise ToolException on REVIEW or BLOCK.
• OpenAI function calling: a guard that runs your function only on ALLOW.
• Anthropic tool use: a check for each tool_use block.
• n8n: a request body for the verify endpoint.
• CrewAI, AutoGen and LlamaIndex: tool wrappers, plus a decorator for any Python function.

Same answers everywhere. Every hookup calls the same service with the same rules. Switching frameworks does not change a verdict.

All hookups in one place, with the full options: Framework hookups in the developer docs.

What every hookup gets back, on one fixed call: ALLOW, REVIEW and BLOCK, measured.