> For the complete documentation index, see [llms.txt](https://intuitem.gitbook.io/ciso-assistant/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://intuitem.gitbook.io/ciso-assistant/configuration/settings/general.md).

# General settings

Instance-wide settings that don't belong to a more specific category. Grouped here by intent — the order in the UI may differ.

## Display and interface

* **Security objective scale** — the labels used for ratings of security objectives (e.g. C/I/A).
* **EBIOS radar configuration** — the max radius and the green / yellow / red zone radii used to draw the EBIOS RM radar chart.
* **Aggregated scenario matrix** — toggles the aggregated view of scenarios on the risk-matrix display.
* **Risk matrix axis options** — `swap axes`, `flip vertical`, custom axis labels. Cosmetic adjustments to how every risk matrix is rendered.
* **Display qualifications as risk categories** — swaps the term *qualification* for *risk category* throughout the interface: field labels on risk scenarios, incidents, feared events, escalation thresholds and quantitative risk scenarios, plus the matching filters and analytics charts. The values themselves (confidentiality, integrity, …) are managed as [terminology](/ciso-assistant/concepts/specialised-modules/terminology.md) either way.

## Language

* **Default language** — fallback locale for users who haven't picked a preference and for system-generated emails. Must be one of the languages enabled in the instance build.

## Money

* **Currency** — the unit used when displaying applied-control costs and quantitative-risk amounts.
* **Daily rate** — default daily cost used when expressing effort in monetary terms.
* **Conversion rate** — write-only field used when changing the currency to convert existing cost amounts in one operation.

> When you change the currency with a conversion rate, the platform sweeps every applied control's cost structure and applies the conversion. Without a conversion rate it just relabels — the numbers stay the same.

## Behaviour

* **Allow self-validation** — whether a user can validate workflows they themselves created. Off by default for separation-of-duty reasons.
* **Show warning on external links** — interstitial prompt before opening links that leave the platform.
* **Enforce MFA** — make multi-factor authentication mandatory for every user account.
* **Allow assignments to entities** — whether requirements and tasks can be assigned to third-party entities (not just internal users).
* **Mapping max depth** — how many frameworks a chained mapping path may span when projecting one framework onto another, counted as nodes in the path (2–5, default 3). A value of 3 allows one pivot — A → B → C. See [transitive inference](/ciso-assistant/concepts/catalog/mappings.md#transitive-inference-pivot-mappings).

## Retention

* **Built-in metrics retention (days)** — how long built-in metric samples are kept before being aged out. Minimum 1.

## Notifications

* **Enable email notifications** — master switch for outbound email notifications. See also [Setting up mailer](/ciso-assistant/installation/post-install-setup/mailer.md).

## AI / LLM provider

These settings drive the optional AI features (chat mode, agentic workflows, RAG over the knowledge base):

* **LLM provider** — which model server the platform calls: `Ollama`, or `OpenAI-compatible (LM Studio, vLLM, llama.cpp...)` for any server speaking the OpenAI chat completions API, hosted or local.
* **Ollama base URL**, **Ollama model**, **Ollama embedding model** — used when the provider is Ollama.
* **API base URL**, **Model name**, **API key** — used when the provider is OpenAI-compatible. The key is write-only and never returned by GET.
* **Embedding backend** — which backend powers semantic search over knowledge. Local sentence transformers or Ollama; hosted providers are not offered here, so selecting one for the LLM leaves embeddings where they are.
* **System prompt** — system prompt prepended to chat-mode conversations.
* **Send temperature to the model** and **Temperature** — whether to pass a sampling temperature, and its value. Some servers reject the parameter, which is why it can be switched off.

To connect a hosted service such as OVHcloud AI Endpoints or OpenRouter, see [Hosted AI providers](/ciso-assistant/ai-and-integrations/ai-providers.md).

{% hint style="warning" %}
**Local or self-hosted LLMs.** The **Ollama URL** and OpenAI **API base URL** are checked when you save them and must resolve to a public address — private, loopback, and internal IPs are rejected to prevent server-side request forgery (SSRF). To point at a local or in-network model (for example Ollama on `localhost`, LM Studio, or an in-cluster endpoint), start the backend with the environment variable `ALLOW_PRIVATE_NETWORK_REQUESTS=True`. (This variable was previously named `WEBHOOK_ALLOW_PRIVATE_IPS`; the old name is no longer recognized.)
{% endhint %}

## Analytics

* **Default custom analytics dashboard** — UUID of the dashboard shown by default on the analytics page.

## Domain-tree audit inheritance

Shown only when the **Domain-tree audit inheritance** feature flag is on (under the **Compliance assessments** group).

* **Domain inheritance strategy** — when the same framework is audited at several levels of the domain tree, decides how a child audit combines results with its parent audits. Options: *No inheritance* (default), *Parent always wins*, *Child always wins*, *Best case (optimistic)*, *Worst case (prudent)*. Scores from a different scale are normalised to the top parent's scale. This one org-wide setting drives the **Combined view** on the [Framework report](/ciso-assistant/features/framework-report.md#combined-view-domain-tree-inheritance) and the inheritance panel in [Advanced Analytics](/ciso-assistant/features/audit-analytics.md). Leaving it on *No inheritance* keeps the feature dormant even with the flag enabled.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://intuitem.gitbook.io/ciso-assistant/configuration/settings/general.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.