Skip to main content
Tuteliq provides machine-readable documentation files so AI agents, coding assistants, and LLM-powered applications can discover and use the API without manual configuration.

​
What agents can do with Tuteliq

Tuteliq covers four capability areas — all accessible via API, MCP tools, or OpenAPI spec:

Content Safety

Detect grooming, bullying, self-harm, radicalisation, and all 9 KOSA harm categories across text, voice, images, and video.

Fraud Detection

Identify social engineering, romance scams, app fraud, and money mule recruitment targeting minors and vulnerable users.

Age & Identity Verification

Verify user age via document analysis, biometric estimation, or both. Full identity verification with face matching and liveness detection for COPPA/KOSA compliance.

Analysis & Reporting

Emotional trend analysis, age-appropriate action plans, and professional incident reports ready for trust & safety teams.

​
Verification endpoints

Verification integrates directly with detection — verify once, then pass the confirmed age_group to all safety endpoints for properly calibrated risk scoring. See the Verification guide for full details.

​
llms.txt

The llms.txt file is a concise, structured summary of Tuteliq — like robots.txt, but for AI. It tells agents what the API does, which endpoints are available, and where to find detailed documentation.
Point your AI assistant to this URL for a quick overview of all capabilities.

​
llms-full.txt

The llms-full.txt file is the complete API reference in plain text — every endpoint, parameter, response field, and example. Ideal for agents that need full context to generate accurate API calls.
If your agent supports a “documentation URL” or “context file” setting, use llms-full.txt for the most comprehensive coverage.

​
OpenAPI Spec

The full OpenAPI 3.1 specification is available at:
This is the same spec that powers the interactive API docs. Use it with any OpenAPI-compatible tool or agent framework.

​
AI Plugin Manifest

For agents that support the OpenAI plugin standard:

​
MCP Server

For AI assistants that support the Model Context Protocol (Claude Desktop, Cursor, Windsurf), Tuteliq exposes all detection tools as MCP tools — no code required. The MCP server also provides built-in data resources (tuteliq://documentation, tuteliq://context-fields, etc.) that agents can read for full parameter references. See the MCP setup guide for configuration instructions.

​
Quick Reference