AI Governance Infrastructure | systemprompt.io

# systemprompt.io > AI governance infrastructure for organisations running agents at scale. A single self-hosted ~50MB Rust binary that authenticates, authorises, rate-limits, logs, and costs every AI interaction. Provider-agnostic across Anthropic Claude, OpenAI, Google Gemini, and local models. MCP-native. PostgreSQL only. Air-gap capable. Source-available under BSL-1.1. systemprompt.io sits between an organisation's AI agents and everything they touch. Every tool call passes through a synchronous four-layer governance pipeline (scope check, secret detection, blocklist, rate limit) before execution. Every decision is logged with a full audit trail emitted as structured JSON ready for Splunk, ELK, Datadog, or any SIEM. Identity flows end-to-end from the user through the agent, through the tool call, to the result and cost, linked by trace_id. The whole system is a compiled Rust binary you run on your own infrastructure; nothing leaves your network. It is a library, not a SaaS product and not a framework. You own the binary and the extension code. It is self-hosted, PostgreSQL is the only dependency, and it can run air-gapped. Two repositories support evaluation: `systempromptio/systemprompt-template` is a clone-and-run template that boots the full binary with demo data in under 10 minutes; `systempromptio/systemprompt-core` is the underlying Rust crate. ## Quick Links - Homepage: http://localhost:8080/?utm_source=llms_txt&utm_medium=ai_agent&utm_campaign=llms_txt - Documentation: http://localhost:8080/documentation?utm_source=llms_txt&utm_medium=ai_agent&utm_campaign=llms_txt - Guides: http://localhost:8080/guides?utm_source=llms_txt&utm_medium=ai_agent&utm_campaign=llms_txt - Evaluation template (clone to try): https://github.com/systempromptio/systemprompt-template - Rust crate (core library): https://github.com/systempromptio/systemprompt-core ## Key Features - **Governance Pipeline**: Synchronous four-layer evaluation on every tool call before execution: