Synthetic Content Detection (5 new tools)

Detect AI-generated content across all modalities with multi-signal forensic analysis:

• detect_synthetic_text — Detect AI-generated text across 10 child-safety categories (synthetic CSAM narratives, deepfake impersonation scripts, AI-generated grooming)
• detect_synthetic_image — 6-signal forensic pipeline: vision AI, EXIF metadata, pixel statistics, C2PA Content Credentials, watermark detection, perceptual hashing
• detect_synthetic_audio — Dual-signal forensics: transcript analysis + mel spectrogram vision + quantitative audio statistics
• detect_synthetic_video — 5-track analysis: per-frame vision, temporal face consistency, lip-sync correlation, spectral audio, transcript detection
• get_synthetic_profile — Account-level 30-day rolling window with trend detection and category distribution

Identity & Age Verification (3 new tools)

Session-based verification — the MCP tool creates a session and provides a URL. The user completes the flow in a hosted web UI (document capture, liveness check, selfie). Poll for results with full document intelligence.

• create_verification_session — Create age or identity verification session, returns URL for the user
• get_verification_session — Poll status with full document intelligence: MRZ validation, PDF417 barcode reading, 45-country document number validation, AI-powered document authenticity, face matching, liveness
• cancel_verification_session — Cancel an active session (no credits consumed)

Credits

Other Changes

• Total tool count: 42 → 50
• Updated credit costs resource
• Requires @tuteliq/sdk@^2.7.1

What's New

Document Analysis Tool

• analyze_document — Analyze PDF documents for safety and compliance concerns
• Extracts text from each page, runs up to 8 detection endpoints in parallel, and returns per-page results with an overall risk assessment
• Supports: unsafe, bullying, grooming, social-engineering, coercive-control, radicalisation, romance-scam, mule-recruitment
• SHA-256 document hashing for chain-of-custody verification
• Zero-retention processing — no document data stored after response
• Dynamic credit pricing: max(3, pages_analyzed × endpoint_count)
• Rich markdown output with flagged pages, extraction summary, and detected threats

Dependencies

• Upgraded @tuteliq/sdk to ^2.6.0

Stats

• 42 tools total across safety, fraud, media, document, analysis, and admin categories

Install / Upgrade

npm install @tuteliq/mcp@3.8.0

Full documentation: https://docs.tuteliq.ai/sdks/mcp

What's New

Added

• country context field — Pass ISO 3166-1 alpha-2 country code (e.g., "GB", "US", "SE") in the context object for geo-localised crisis helpline data.
• Updated context fields documentation in MCP resources.

Changed

• Bumped @tuteliq/sdk dependency to ^2.5.0.

Install

npm install @tuteliq/mcp@3.7.0

Or in Claude Desktop / Cursor config:

{
  "mcpServers": {
    "tuteliq": {
      "command": "npx",
      "args": ["-y", "@tuteliq/mcp@3.7.0"],
      "env": { "TUTELIQ_API_KEY": "your-api-key" }
    }
  }
}

What's New

MCP Data Resources for AI Agents

Agents consuming the Tuteliq MCP server can now read documentation resources directly within the MCP session — no need to navigate external docs.

New resources:

Why This Matters

AI agents (Claude, Cursor, Windsurf) can now discover the full parameter surface — context fields, support_threshold values, analyse_multi endpoint IDs — without trial and error. This directly addresses the documentation gap identified in user testing.

Full Changelog

v3.5.1...v3.6.0

Documentation

• Add support_threshold parameter documentation
• Add sender_trust effect on scoring
• Add analyse_multi accepted endpoint values table
• Add complete context fields reference

What's Changed

• Wire support_threshold parameter through all detection MCP tools (bullying, grooming, unsafe)
• Show confidence in analyze tool output
• Show support resources from unsafe sub-result in analyze tool
• Bump @tuteliq/sdk dependency to ^2.4.0

What's New

Fraud & Harm Detection (8 new tools)

• detect_social_engineering — Pretexting, urgency fabrication, authority impersonation
• detect_app_fraud — Fake investment platforms, phishing apps, subscription traps
• detect_romance_scam — Love-bombing, financial requests, identity deception
• detect_mule_recruitment — Easy-money offers, bank account sharing tactics
• detect_gambling_harm — Chasing losses, concealment, gambling-related distress
• detect_coercive_control — Isolation, financial control, monitoring, threats
• detect_vulnerability_exploitation — Targeting elderly, disabled, or financially distressed individuals
• detect_radicalisation — Extremist rhetoric, us-vs-them framing, ideological grooming

Crisis Support Card

• Country-specific crisis helplines displayed in the MCP widget when high/critical threats are detected
• Warm, category-aware messages (self-harm, bullying, grooming, etc.)
• Phone numbers copy to clipboard on click with confirmation feedback
• Resource links open in system browser via MCP Apps app.openLink() API

Language Support

• 27 languages now supported with auto-detection
• English (stable) + 26 beta languages including Spanish, Portuguese, French, German, Italian, Swedish, Norwegian, Finnish, Ukrainian, and more

UI Improvements

• New components: SupportCard, BrandedLoader, FooterLinks, UpsellBanner
• All widget links now work correctly in Claude Desktop via app.openLink()

Security

• Fixed Hono prototype pollution vulnerability (GHSA-v8w9-8mx6-g223)
• Removed internal infrastructure references from tool descriptions

Other

• Updated README with all 41 MCP tools, fraud section, languages table
• Verification tools documented as API & SDK only
• Added .claude/ to .gitignore

Full Changelog: v3.2.0...v3.3.4

What's Changed

• Added security tooling: Dependabot, CodeQL analysis, and npm audit CI workflow
• Version bump to 3.2.0

What's Changed

Dependencies

• Bump @modelcontextprotocol/sdk from ^1.12.0 to ^1.27.0
• Bump @modelcontextprotocol/ext-apps from ^1.0.0 to ^1.1.0
• Bump @tuteliq/sdk from ^2.2.2 to ^2.3.1

Security

• Fixed 4 security vulnerabilities (hono, @hono/node-server, ajv, rollup)

MCP Apps with Interactive UI Widgets

Breaking Changes

• Modular tool architecture (monolithic src/index.ts split into src/tools/)
• New SSE transport support alongside stdio
• Version bump to 3.0.0

New Features

• Interactive UI widgets for rich result visualization (detection, multi-analysis, emotions, media, action plan, report)
• Fraud detection tools (social engineering, app fraud, romance scam, mule recruitment, gambling harm, coercive control, vulnerability exploitation, radicalisation)
• Multi-endpoint analysis (analyseMulti)
• Video analysis (analyzeVideo)
• Dockerfile for containerized deployment

Fixes

• Updated package-lock.json to resolve @tuteliq/sdk@2.2.1 (was pinned to 2.1.0)