An open-source CLI + API to classify AI systems under the EU AI Act into risk tiers — prohibited, high, limited, or minimal-risk using a versioned JSON ruleset, not LLM. Describe your AI system (purpose, data types, deployment context, autonomy, who it affects) and get back a risk tier with clause references, a plain-language rationale, and recommended controls. Built for engg teams. Features: Deterministic classification with, auditable ruleset Offline-first — no data leaves your machine
No reviews yetBe the first to leave a review for riskline
Maker
📌
Hey Product Hunt! I built RiskLine because most "AI governance" tools want you inside a portal — but engineering teams already live in CI and the terminal.
Classification that can't run offline, can't be diffed, and can't explain which article fired is hard to trust. RiskLine is a small, embeddable engine that gives you a risk tier with clause references you can actually verify.
It's pre-1.0 and I'm actively collecting misclassification reports to validate the ruleset. Would love feedback from anyone dealing with EU AI Act compliance.
go install github.com/new-world-coder/riskl...
Topics: Open Source, Developer Tools, Artificial Intelligence, Compliance, Legal Tech
Thumbnail/Logo: Use the repo's social card or a simple icon with "RiskLine" text.
Links:
GitHub: https://github.com/new-world-cod...
Methodology write-up: https://github.com/new-world-cod...