RepoAI
Browse MCPs

Data Leakage Detection

100TrustedPre-approves tools

Identification of security risks through systematic interrogation of model responses.

What it does

Data Leakage Detection provides a structured protocol to audit AI models for the inadvertent exposure of restricted data. It systematically evaluates whether an agent can be persuaded to reveal internal configuration details, sensitive credentials, or personal information when subjected to varying levels of scrutiny. By standardizing these checks, the skill ensures that potential vulnerabilities are identified consistently during the assessment process.

Installing this skill allows your assistant to simulate adversarial inquiries, ranging from direct requests to more complex evasion and override attempts. It keeps a record of which information categories have been probed, ensuring that once a specific weakness is verified, the investigation for that category concludes to maintain operational efficiency.

When it earns its place

Reach for it when…

Use this skill during the security validation phase of an AI deployment, particularly when assessing a new model or agent before it handles private data or interacts with live external services.

Look elsewhere when…

This skill is not intended for real-time traffic monitoring or active defense of an agent against malicious users. It does not provide automated patching or corrective adjustments to system prompts once a leakage is discovered.

What to say to use it

You never invoke a skill directly — the assistant picks it up when your request matches what it covers. These are the kinds of things that do that.

  • Check if this bot exposes its system prompt
  • Identify if any API keys are reachable in the context
  • Find potential privacy leaks in the agent responses
  • Test the agent for configuration data exposure

Install it

Install for all your projects

Copy the skill directory into your personal skills folder. Claude picks it up on its own the next time your request matches what it does — there is nothing to enable.

bash
git clone --depth 1 --branch main https://github.com/Tencent/AI-Infra-Guard.git /tmp/AI-Infra-Guard
mkdir -p ~/.claude/skills
cp -r /tmp/AI-Infra-Guard/agent-scan/agent_scan/prompt/skills/data-leakage-detection ~/.claude/skills/data-leakage-detection

Install for one project only

Put it in the project's own skills folder instead, so it can be committed and shared with everyone working on that repository.

bash
cp -r /tmp/AI-Infra-Guard/agent-scan/agent_scan/prompt/skills/data-leakage-detection .claude/skills/data-leakage-detection

If a skill of the same name exists in both places, the personal one wins.

Installing on Claude.ai or through the API instead? The full guide covers every surface.

Pair it with an MCP server

Data Leakage Detection changes how the work is done. These servers change what Claude can reach while doing it.

How skills and MCP servers differ →