Help Center / AI Assistant / Configuration

Configuration

The AI Assistant lets you analyze risk data through natural language. It is off until you supply a provider and an API key. MTC Skopos never proxies these calls: your client talks directly to the provider you configure, with your key.

Provider setup

Provider typeTypical modelsNotes
AnthropicClaude Sonnet, Claude OpusThe default
OpenAI CompatibleGPT-class models, and any OpenAI-compatible endpoint including local serversSet a base URL for anything other than the public OpenAI API
Azure OpenAIAzure-hosted deploymentsNeeds a base URL, API version, and auth header
  1. Go to Settings > AI Assistant
  2. Select your Provider Type
  3. Enter your API Key
  4. Choose a Model. The shipped default is claude-sonnet-4-20250514
  5. Set a Base URL if you use a self-hosted, proxy, or Azure endpoint

Azure OpenAI

Azure deployments need three extra fields:

FieldDefault
Base URLnone, required
API Version2024-10-21
Auth Headerapi-key

Model settings

SettingDefaultWhat it does
Temperature0.3Response variability. 0.0 is near-deterministic, 1.0 is creative. Low values suit audit work
Max Output Tokens16384Ceiling on the length of a single response
Response LanguageEnglishLanguage the model answers in
Context Window200,000Total conversation budget before compaction kicks in

Two further settings govern long conversations: a reserve of 30,000 tokens is held back for the response, and 40,000 tokens of recent turns are kept verbatim when older turns are compacted. You rarely need to change either.

Tool limits

These caps decide how much data any single tool call may return. They exist because analysis results are frequently far larger than a context window.

SettingDefaultCeiling
Batch size1025
Search results10100
Entities per risk20100
Overview top N10100
Aggregation groups20100
Detail items50100
Flat recommendations30100

The ceiling is a hard cap in the application. Asking the model for "all users" does not lift it, so a broad question against a large result returns a truncated view rather than an error. Where completeness matters, scope the analysis narrowly first and ask the assistant about the smaller result. See Scoping Your Analysis.

Lower values produce faster, cheaper answers. Raise them when you need breadth and accept the token cost.

Updated on: 2026-08-02

« Back to AI Assistant