Configure and enforce content safety guardrails for TrueFoundry AI Gateway.
npx skills add https://github.com/passionseed/ps_app --skill guardrails-passionseed
Please help me install this Agent Skill. Skill: guardrails Source: https://github.com/passionseed/ps_app/tree/main/.agents/skills/truefoundry-guardrails Command: npx skills add https://github.com/passionseed/ps_app --skill guardrails-passionseed
Guardrails provide configurable safety controls for TrueFoundry AI Gateway, including PII filtering, content moderation, prompt-injection detection, and custom rules, to reduce risk in model and tool interactions.
Install and configure a guardrail by creating a Guardrail Config Group with provider integrations, then attach it to a Gateway Guardrails Config and apply to a gateway.
High-intent search queries and answers about installing and using this skill.
To add content safety guardrails to an AI Gateway, you configure a Guardrail Config Group with provider integrations, then attach it to a Gateway Guardrails Config and apply it to the gateway deployment to filter PII and enforce policy compliance.
Yes, you can filter PII and detect prompt injections across LLM inputs and outputs and MCP tool calls by configuring content moderation, secret detection, and custom rules within the gateway guardrails config.
Configuring policy compliance for MCP tool calls involves creating a Guardrail Config Group with the necessary provider integrations and attaching it to a Gateway Guardrails Config to enforce custom rules during tool interactions.
Yes, you need to deploy guardrails servers using the official template before attaching the config to ensure gateway compatibility and that provider references resolve to valid config integrations.
Custom rule enforcement for LLM outputs depends on valid provider references resolving to existing config integrations, and you must deploy the guardrails server using the official template to maintain gateway compatibility and prevent unsafe outputs.
truefoundryConfigure content safety guardrails for AI Gateway endpoints.
truefoundryConfigure content safety guardrails for the TrueFoundry AI Gateway.
truefoundryConfigure TrueFoundry AI Gateway with OpenAI-compatible endpoints, auth, routing, and rate limiting.
LauraFlorentinInspect inputs and outputs to enforce safety and policy compliance.
