Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add mukul975/Privacy-Data-Protection-Skills --skill llm-output-privacy-riskgit clone --depth 1 https://github.com/mukul975/Privacy-Data-Protection-SkillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mukul975/privacy-data-protection-skills/llm-output-privacy-risk)<a href="https://agentmods.dev/skills/mukul975/privacy-data-protection-skills/llm-output-privacy-risk"><img src="https://agentmods.dev/badge/skills/mukul975/privacy-data-protection-skills/llm-output-privacy-risk/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mukul975/privacy-data-protection-skills/llm-output-privacy-risk"><img src="https://agentmods.dev/badge/skills/mukul975/privacy-data-protection-skills/llm-output-privacy-risk.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00076 | $0.02190 |
| Opus 5 | $0.00038 | $0.01095 |
| Sonnet 5 | $0.00015 | $0.00438 |
| Haiku 4.5 | $0.00008 | $0.00219 |
Grade B, and why
llm-output-privacy-risk scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 13d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Instruction-override phrasingmediumPrompt injection
Text telling the model to disregard its earlier instructions or safety rules is the shape of a prompt injection, whoever wrote it.
Adversarial prompts can cause models to bypass safety instructions and output training data, system prompts, or information about other users' conversations. Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
How it starts
The opening of the file, as written. The whole thing — 170 lines — stays where its author put it; the contents beside it link to each section on GitHub.
LLM Output Privacy Risk Assessment
Overview
Large language models (LLMs) present unique privacy risks that go beyond traditional ML systems. Because LLMs are trained on massive corpora that may contain personal data, they can memorise and reproduce verbatim training data — including names, email addresses, phone numbers, and other PII. Additionally, LLMs can hallucinate plausible but false personal data, creating defamation and accuracy risks. Prompt injection attacks can bypass safety guardrails to extract training data or system prompts containing confidential information. This skill provides a structured framework for assessing, mitigating, and monitoring privacy risks in LLM-generated outputs at Cerebrum AI Labs.
LLM Output Privacy Risk Categories
Risk 1: Training Data Memorisation
LLMs memorise training data, particularly sequences that appear multiple times or are distinctive. Extractable memorisation occurs when a model, given a prefix, completes the text with verbatim training data.
| Factor | Impact on Memorisation Risk |
|---|---|
| Model size | Larger models memorise more (Carlini et al., 2023) |
| Data duplication | Repeated sequences are memorised at higher rates |
| Training epochs | More passes over data increase memorisation |
| Data distinctiveness | Unique sequences (names, numbers) are more extractable |
| Temperature at inference | Lower temperature increases verbatim reproduction |
Regulatory concern: If a model reproduces personal data from training, this constitutes processing under GDPR Art. 4(2). The data subject has not consented to this output, and the controller must have a lawful basis for the disclosure.
Risk 2: PII Leakage in Generated Text
Even without verbatim memorisation, models can combine partial information to produce outputs containing personal data — email patterns, phone number formats with real area codes, or names associated with specific contexts.
| Leakage Type | Example | Detection Method |
|---|---|---|
| Verbatim reproduction | Model outputs exact email address from training data | Exact match against known training PII |
| Recombination | Model combines real first name + real surname from different records | Named entity recognition + cross-reference |
| Pattern completion | Model generates plausible phone number matching real format | Regex + validation against real registries |
| Contextual leakage | Model associates real person with correct employer when prompted | Entity relationship extraction |
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 13d ago First seen · 170 lines · 76 tokens per session scan B 3c8cd1715994
llm-output-privacy-risk is a skill published in the GitHub repository mukul975/Privacy-Data-Protection-Skills (272 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 76 tokens to every session and 2,190 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 1 finding (instruction-override phrasing). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
privacy-compliance
Comprehensive global privacy compliance agent skill covering GDPR, CCPA/CPRA, HIPAA Privacy Rule, EU AI Act, LGPD, cross-border data transfer mechanisms (SCCs, BCRs, EU-US DPF), PII identification and classification, data minimization, consent management, privacy-by-design patterns, DPIA workflows, data subject access…
fedramp
Expert guidance for FedRAMP certification and compliance under CR26 (FedRAMP Consolidated Rules for 2026). Use this skill whenever a user asks about FedRAMP authorization, ATO (Authority to Operate), cloud security for federal government, NIST SP 800-53 controls, CSP compliance, or any of the core FedRAMP document…
gdpr-compliance
Expert GDPR compliance assistant covering all four core workflows: (1) auditing code and systems for GDPR violations, (2) drafting GDPR-compliant documents such as privacy policies, Data Processing Agreements (DPAs), and consent notices, (3) answering GDPR compliance questions with authoritative article citations, and…
iso42001
Expert ISO 42001 AI Management System (AIMS) compliance advisor. Use this skill whenever a user asks about ISO/IEC 42001:2023, AI governance, AI management systems, AI risk assessment, AI system impact assessment, Annex A controls for AI, Statement of Applicability for AI systems, AI policy, responsible AI, AI…
eu-cra
Expert EU Cyber Resilience Act (CRA) advisor for Regulation (EU) 2024/2847 — mandatory cybersecurity and vulnerability handling requirements for all products with digital elements (PDEs) sold in the EU. Use this skill for gap analysis, product classification (Default / Class I / Class II), conformity assessment route…
nist-800-53
NIST SP 800-53 Rev 5 compliance advisor — all 20 control families (AC, AT, AU, CA, CM, CP, IA, IR, MA, MP, PE, PL, PM, PS, PT, RA, SA, SC, SI, SR), Low/Moderate/High baseline selection, FIPS 199/200 system categorization, control tailoring and overlays, privacy controls (PT family), supply chain risk management (SR…