perplexity-data-handling

0
0
Source

Implement Perplexity PII handling, data retention, and GDPR/CCPA compliance patterns. Use when handling sensitive data, implementing data redaction, configuring retention policies, or ensuring compliance with privacy regulations for Perplexity integrations. Trigger with phrases like "perplexity data", "perplexity PII", "perplexity GDPR", "perplexity data retention", "perplexity privacy", "perplexity CCPA".

Install

mkdir -p .claude/skills/perplexity-data-handling && curl -L -o skill.zip "https://mcp.directory/api/skills/download/6012" && unzip -o skill.zip -d .claude/skills/perplexity-data-handling && rm skill.zip

Installs to .claude/skills/perplexity-data-handling

About this skill

Perplexity Data Handling

Overview

Manage data flowing through Perplexity Sonar API. Critical concern: queries are sent to Perplexity for web search, so any PII in queries is exposed to external infrastructure. Responses contain citations (third-party URLs) that must be validated before displaying to users.

Data Flow

User Input → Query Sanitization → Perplexity API → Response Parsing
                                                         │
                                           ┌─────────────┼──────────────┐
                                           │             │              │
                                      Answer Text    Citations    Search Results
                                           │             │              │
                                      Format &      Validate &    Store for
                                      Display       Deduplicate   Analytics

Prerequisites

  • Perplexity API key configured
  • Understanding of PII regulations (GDPR/CCPA)
  • Cache storage (Redis or in-memory)

Instructions

Step 1: Query Sanitization

function sanitizeQuery(query: string): { clean: string; redacted: boolean } {
  let clean = query;
  let redacted = false;

  const patterns: Array<[RegExp, string]> = [
    [/\b[\w.+-]+@[\w-]+\.[\w.]+\b/g, "[email]"],
    [/\b\d{3}[-.]?\d{3}[-.]?\d{4}\b/g, "[phone]"],
    [/\b\d{3}-\d{2}-\d{4}\b/g, "[ssn]"],
    [/\b\d{4}[\s-]?\d{4}[\s-]?\d{4}[\s-]?\d{4}\b/g, "[card]"],
    [/\b(pplx-|sk-|pk_|sk_live_)\w{20,}\b/g, "[token]"],
    [/\b(user|customer|account)\s*#?\s*\d+\b/gi, "[id]"],
  ];

  for (const [pattern, replacement] of patterns) {
    if (pattern.test(clean)) {
      clean = clean.replace(pattern, replacement);
      redacted = true;
    }
  }

  return { clean, redacted };
}

async function safeSearch(rawQuery: string) {
  const { clean, redacted } = sanitizeQuery(rawQuery);
  if (redacted) {
    console.warn("[Data] PII redacted from Perplexity query");
  }

  return perplexity.chat.completions.create({
    model: "sonar",
    messages: [{ role: "user", content: clean }],
  });
}

Step 2: Citation Validation

interface ValidatedCitation {
  url: string;
  domain: string;
  valid: boolean;
  index: number;
}

function validateCitations(citations: string[]): ValidatedCitation[] {
  return citations.map((url, i) => {
    try {
      const parsed = new URL(url);
      return {
        url: url.replace(/[.,;:]+$/, ""),
        domain: parsed.hostname,
        valid: ["http:", "https:"].includes(parsed.protocol),
        index: i + 1,
      };
    } catch {
      return { url, domain: "unknown", valid: false, index: i + 1 };
    }
  });
}

function deduplicateCitations(citations: ValidatedCitation[]): ValidatedCitation[] {
  const seen = new Set<string>();
  return citations.filter((c) => {
    const normalized = c.url.split("?")[0].replace(/\/$/, "");
    if (seen.has(normalized)) return false;
    seen.add(normalized);
    return true;
  });
}

// Replace [1] markers with linked citations
function renderCitations(answer: string, citations: ValidatedCitation[]): string {
  let rendered = answer;
  for (const c of citations.filter((c) => c.valid)) {
    rendered = rendered.replaceAll(`[${c.index}]`, `[${c.index}](${c.url})`);
  }
  return rendered;
}

Step 3: Result Caching with Freshness Policy

import { LRUCache } from "lru-cache";
import { createHash } from "crypto";

interface CachedResult {
  answer: string;
  citations: ValidatedCitation[];
  cachedAt: number;
  model: string;
}

const CACHE_TTL: Record<string, number> = {
  news: 30 * 60_000,       // 30 min for breaking/current events
  research: 4 * 3600_000,  // 4 hours for research topics
  factual: 24 * 3600_000,  // 24 hours for stable facts
  default: 1 * 3600_000,   // 1 hour default
};

const resultCache = new LRUCache<string, CachedResult>({ max: 500 });

function detectQueryType(query: string): keyof typeof CACHE_TTL {
  if (/\b(latest|today|breaking|recent|this week)\b/i.test(query)) return "news";
  if (/\b(research|study|paper|analysis|compare)\b/i.test(query)) return "research";
  if (/\b(what is|define|how does|who is)\b/i.test(query)) return "factual";
  return "default";
}

async function cachedSearch(query: string, model = "sonar") {
  const hash = createHash("sha256")
    .update(`${model}:${query.toLowerCase().trim()}`)
    .digest("hex");

  const cached = resultCache.get(hash);
  if (cached) return { ...cached, fromCache: true };

  const response = await safeSearch(query);
  const rawCitations = (response as any).citations || [];
  const citations = deduplicateCitations(validateCitations(rawCitations));
  const queryType = detectQueryType(query);

  const entry: CachedResult = {
    answer: response.choices[0].message.content || "",
    citations,
    cachedAt: Date.now(),
    model: response.model,
  };

  resultCache.set(hash, entry, { ttl: CACHE_TTL[queryType] });
  return { ...entry, fromCache: false };
}

Step 4: Conversation Context Management

import OpenAI from "openai";

type Message = OpenAI.ChatCompletionMessageParam;

class SearchContext {
  private messages: Message[] = [];
  private readonly maxMessages = 10;
  private readonly maxEstimatedTokens = 8000;

  constructor(systemPrompt?: string) {
    if (systemPrompt) {
      this.messages.push({ role: "system", content: systemPrompt });
    }
  }

  addUserMessage(content: string) {
    this.messages.push({ role: "user", content });
    this.trim();
  }

  addAssistantMessage(content: string) {
    this.messages.push({ role: "assistant", content });
    this.trim();
  }

  getMessages(): Message[] {
    return [...this.messages];
  }

  private trim() {
    // Keep system prompt + last N messages
    while (this.messages.length > this.maxMessages) {
      const systemIdx = this.messages[0].role === "system" ? 1 : 0;
      this.messages.splice(systemIdx, 1);
    }

    // Trim if estimated tokens too high
    while (this.estimateTokens() > this.maxEstimatedTokens && this.messages.length > 2) {
      const systemIdx = this.messages[0].role === "system" ? 1 : 0;
      this.messages.splice(systemIdx, 1);
    }
  }

  private estimateTokens(): number {
    return this.messages.reduce(
      (sum, m) => sum + Math.ceil(String(m.content).length / 4),
      0
    );
  }

  clear() {
    const system = this.messages.find((m) => m.role === "system");
    this.messages = system ? [system] : [];
  }
}

Error Handling

IssueCauseSolution
PII in search queryUser entered personal dataApply sanitizeQuery before API call
Broken citation URLsSource page moved/deletedValidate URLs, filter invalid ones
Stale cached resultsTTL too long for newsUse query-type-aware TTL
Context overflowToo many conversation turnsAutomatic trimming in SearchContext
Duplicate citationsSame source cited multiple timesDeduplicate by normalized URL

Output

  • Query sanitization stripping PII before API calls
  • Citation validation and deduplication
  • Cache with query-type-aware TTL
  • Conversation context with automatic trimming

Resources

Next Steps

For access control, see perplexity-enterprise-rbac.

svg-icon-generator

jeremylongshore

Svg Icon Generator - Auto-activating skill for Visual Content. Triggers on: svg icon generator, svg icon generator Part of the Visual Content skill category.

6814

d2-diagram-creator

jeremylongshore

D2 Diagram Creator - Auto-activating skill for Visual Content. Triggers on: d2 diagram creator, d2 diagram creator Part of the Visual Content skill category.

2412

performing-penetration-testing

jeremylongshore

This skill enables automated penetration testing of web applications. It uses the penetration-tester plugin to identify vulnerabilities, including OWASP Top 10 threats, and suggests exploitation techniques. Use this skill when the user requests a "penetration test", "pentest", "vulnerability assessment", or asks to "exploit" a web application. It provides comprehensive reporting on identified security flaws.

379

designing-database-schemas

jeremylongshore

Design and visualize efficient database schemas, normalize data, map relationships, and generate ERD diagrams and SQL statements.

978

performing-security-audits

jeremylongshore

This skill allows Claude to conduct comprehensive security audits of code, infrastructure, and configurations. It leverages various tools within the security-pro-pack plugin, including vulnerability scanning, compliance checking, cryptography review, and infrastructure security analysis. Use this skill when a user requests a "security audit," "vulnerability assessment," "compliance review," or any task involving identifying and mitigating security risks. It helps to ensure code and systems adhere to security best practices and compliance standards.

86

django-view-generator

jeremylongshore

Generate django view generator operations. Auto-activating skill for Backend Development. Triggers on: django view generator, django view generator Part of the Backend Development skill category. Use when working with django view generator functionality. Trigger with phrases like "django view generator", "django generator", "django".

15

You might also like

flutter-development

aj-geddes

Build beautiful cross-platform mobile apps with Flutter and Dart. Covers widgets, state management with Provider/BLoC, navigation, API integration, and material design.

643969

drawio-diagrams-enhanced

jgtolentino

Create professional draw.io (diagrams.net) diagrams in XML format (.drawio files) with integrated PMP/PMBOK methodologies, extensive visual asset libraries, and industry-standard professional templates. Use this skill when users ask to create flowcharts, swimlane diagrams, cross-functional flowcharts, org charts, network diagrams, UML diagrams, BPMN, project management diagrams (WBS, Gantt, PERT, RACI), risk matrices, stakeholder maps, or any other visual diagram in draw.io format. This skill includes access to custom shape libraries for icons, clipart, and professional symbols.

591705

ui-ux-pro-max

nextlevelbuilder

"UI/UX design intelligence. 50 styles, 21 palettes, 50 font pairings, 20 charts, 8 stacks (React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind). Actions: plan, build, create, design, implement, review, fix, improve, optimize, enhance, refactor, check UI/UX code. Projects: website, landing page, dashboard, admin panel, e-commerce, SaaS, portfolio, blog, mobile app, .html, .tsx, .vue, .svelte. Elements: button, modal, navbar, sidebar, card, table, form, chart. Styles: glassmorphism, claymorphism, minimalism, brutalism, neumorphism, bento grid, dark mode, responsive, skeuomorphism, flat design. Topics: color palette, accessibility, animation, layout, typography, font pairing, spacing, hover, shadow, gradient."

318398

godot

bfollington

This skill should be used when working on Godot Engine projects. It provides specialized knowledge of Godot's file formats (.gd, .tscn, .tres), architecture patterns (component-based, signal-driven, resource-based), common pitfalls, validation tools, code templates, and CLI workflows. The `godot` command is available for running the game, validating scripts, importing resources, and exporting builds. Use this skill for tasks involving Godot game development, debugging scene/resource files, implementing game systems, or creating new Godot components.

339397

nano-banana-pro

garg-aayush

Generate and edit images using Google's Nano Banana Pro (Gemini 3 Pro Image) API. Use when the user asks to generate, create, edit, modify, change, alter, or update images. Also use when user references an existing image file and asks to modify it in any way (e.g., "modify this image", "change the background", "replace X with Y"). Supports both text-to-image generation and image-to-image editing with configurable resolution (1K default, 2K, or 4K for high resolution). DO NOT read the image file first - use this skill directly with the --input-image parameter.

451339

fastapi-templates

wshobson

Create production-ready FastAPI projects with async patterns, dependency injection, and comprehensive error handling. Use when building new FastAPI applications or setting up backend API projects.

304231

Stay ahead of the MCP ecosystem

Get weekly updates on new skills and servers.