Run ValGuard with the tools you already use

Keep Agents SDK workflows. Add ValGuard as the OpenAI-compatible base URL so input and output policies stay auditable across models.

Insert ValGuard between LangGraph nodes so a failed schema or policy check cannot become the next node's input.

Point ChatOpenAI (or any OpenAI-compatible LangChain chat model) at ValGuard. Same agent slug and rule packs across Python services.

Configure the OpenAI-compatible LLM client in LlamaIndex with ValGuard as base URL so retrieval-augmented answers pass policy before tools run.

Point each CrewAI agent LLM at ValGuard. Task guardrails in CrewAI stay in-process; the proxy shares packs across crews and other languages.

Use createOpenAI with ValGuard base URL and X-VG-Agent. Block and re-ask rules buffer the stream before the UI sees tokens.

Set ValGuard as api_base for a LiteLLM model alias so routing and fallbacks sit upstream while rules run on every completion.

Register ValGuard as a custom OpenAI-compatible host in Portkey. Disable cache on paths that must hit validators every time.

Use openrouter/ model prefixes through ValGuard Provider Vault, or point a client at ValGuard when OpenRouter is your upstream.

Enforce packs on the ValGuard path. Attach X-Request-Id and validation status to Langfuse generation metadata so blocks show up in the same timeline.

Point the Vapi assistant custom LLM URL at ValGuard so spoken-agent prompts and completions pass the same rule packs as chat.

Configure Retell to call ValGuard as the OpenAI-compatible LLM endpoint. Latency includes model time plus any block or re-ask buffering.

Point n8n LLM credentials at ValGuard or call the chat-completions endpoint from an HTTP Request node. Branch on validation results.

Route MCP host model calls through ValGuard. Tool side effects still need explicit gates — see the MCP security guide.