GitHub - yashmahajan10/llm-differential-privacy-gateway: Noisegate: a differential privacy gateway that lets an untrusted LLM agent query sensitive data over MCP (Model Context Protocol), with a formal guarantee no individual's record can leak even if the agent is adversarial - enforcement lives in trusted code below the model, validated by a runnable attack gallery.

G github.com ↗

▲ 19 points • 0 comments • by yashmahajan10 • 4w ago • HN discussion ↗

Pangram verdict · v3.3

We believe this text is mainly AI, with some human-written content.

92 %

AI likelihood · overall

2% human-written 98% AI-generated

SEGMENTS · HUMAN 0 of 1

SEGMENTS · AI 0 of 1

WORD COUNT 43

PEAK AI % 39% · §1

Analyzed

Jul 30

backend: pangram/v3.3

Segments scanned

1 windows

avg 43 words each

Distribution

2 / 98%

human / AI fraction

Verdict

Pangram v3.3

Article text · 43 words · 1 segments analyzed

Human AI-generated

§1 Mixed · 39%

Give an AI agent query access to sensitive data, with a mathematical guarantee that no individual's record can leak — even if the agent is wrong, manipulated, or adversarial.

A recorded Claude Desktop session (replies trimmed; the chart cards are the session's own).