OSGi Configurations

This is an automatically generated overview.

Composum AI Approximate Markdown Service Configuration (slingbase)

Configuration for the Approximate Markdown Service used to get a text representation of a page or component for use with the AI.

id	name	type	description
urlSourceWhitelist	URL Source Whitelist Regex	String[]	Only if using URLs as external source: Whitelist for URLs that can be read and turned into markdown. If not set, reading URLs is turned off. For security reasons you might want to prevent local addresses to be contacted. To allow everything you might use https?://.* , but make sure you have a good blacklist in that case.
urlSourceBlacklist	URL Source Blacklist Regex	String[]	Only if using URLs as external source: Blacklist for URLs that can be read and turned into markdown. Has precendence over whitelist.
textAttributes	Text Attributes	String[]	List of attributes that are treated as text and converted to markdown. If not present, no attributes are treated as text.
labelledAttributePatternAllow	Labeled Attribute Pattern Allow	String[]	Regular expressions for attributes that are output with a label. If not present, none will be output except the text attributes.
labelledAttributePatternDeny	Labeled Attribute Pattern Deny	String[]	Regular expressions for attributes that are not output with a label. Takes precedence over the corresponding allow regexp list.
labelledAttributeOrder	Labelled Attribute Order	String[]	List of labelled attributes that come first if they are present, in the given order.

Composum AI Autotranslate Configuration (aem/core)

Configuration of the automatic translation of AEM pages. The OSGI configuration is only used if no Sling CAConfig configuration is found.

id	name	type	default value	description
pocUiEnabled	Debugging UI	boolean	false	Enable the Autotranslate debugging UI at at /apps/composum-ai/components/autotranslate/list/list.html, normally disabled. Only read from OSGI configuration.
disabled	Disable	boolean	false	Disable the Autotranslate service.
deniedResourceTypes	Denied Resource Types	String[]	[]	Regexes for denied Resource Types - if the sling:resourceType matches that, then no attributes or child nodes are touched by the automatic translation.
allowedAttributeRegexes	Allowed Additional Attributes	String[]	[]	Matches for Attributes that are explicitly allowed to be translated, in addition to standard attributes and heuristically recognized attributes. The heuristics is that the value has to have letters and whitespaces. Syntax: regular expressions that match resource type % attribute name - e.g. myapp/component/html%markup
deniedAttributesRegexes	Denied Attributes	String[]	[]	Attributes that are explicitly denied to be translated. Can be used to override the heuristics / standard attributes. Same syntax as allowed attributes.
ignoreAssetsExceptContentFragments	Ignore Assets except Content Fragments	boolean	true	If true, assets are ignored for translation, except for content fragments. (Otherwise the translator would translate metadata of images and videos.)
useHighIntelligenceModel	Use High Intelligence Model	boolean	true	If true, the translator will use the ‘high-intelligence model’ (see OpenAI config) for translation. Default: true.
includeFullPageInRetranslation	Include Full Page during Retranslation	boolean	true	If true we do not only provide changed texts to the AI during re-translating a page with some changes, but give the entire page to provide better context. That is a bit slower and a bit more expensive, but likely improves the result.
includeExistingTranslationsInRetranslation	Include Existing Translations in Retranslation	boolean	true	If true, when retranslating a page with some changes we provide the existing translations of that page to the AI as well as additional context with examples. That is a bit slower and a bit more expensive, but likely improves the result.

Composum AI OpenAI Configuration (backend/base)

Provides rather low level access to the GPT chat completion - use the other services for more specific services.

id	name	type	default value	description
disabled	Disable	boolean	false	Disable the GPT Chat Completion Service
chatCompletionUrl	URL of the chat completion service	String		Optional, if not OpenAI's default https://api.openai.com/v1/chat/completions
openAiApiKey	OpenAI API key	String		OpenAI API key from https://platform.openai.com/. If not given, we check the key file, the environment Variable OPENAI_API_KEY, and the system property openai.api.key .
openAiOrganizationId	OpenAI Organization ID	String		Optionally, OpenAI Organization ID from https://platform.openai.com/account/organization .
openAiApiKeyFile	OpenAI API key file	String		Key File containing the API key, as an alternative to Open AI Key configuration and the variants described there.
defaultModel	Default model	String	gpt-4o-mini	Default model to use for the chat completion. The default if not set is gpt-4o-mini. Please consider the varying prices https://openai.com/pricing .
highIntelligenceModel	High intelligence model	String	gpt-4o	The model that is used for requests that need more reasoning performance. The default if not set is gpt-4o. Please consider the varying prices https://openai.com/pricing .
imageModel	Vision model	String	gpt-4o	Optional, a model that is used if an image is given as input, e.g. gpt-4o. If not given, image recognition is rejected.
temperature	Temperature	String		Optional temperature setting that determines variability and creativity as a floating point between 0.0 and 1.0
seed	seed	String		If specified, OpenAI will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.
maximumTokensPerRequest	Maximum Tokens per Request	int	50000	If > 0 limit to the maximum number of tokens per request. That's about a twice the word count. Caution: Compare with the pricing - on GPT-4 models a thousand tokens might cost $0.01 or more.
maximumTokensPerResponse	Maximum output tokens per request	int	4096	Maximum number of tokens to return in the response. Must not exceed the capabilities of the model - as of 10/03/24 this is 4096 for most OpenAI models - which is the default, so no need to set that.
connectionTimeout	Connection timeout in seconds	int	30	Default 30
requestTimeout	Request timeout in seconds	int	300	Default 300
requestsPerMinute	Maximum requests per minute	int	100	Maximum count of requests to ChatGPT per minute - from the second half there will be a slowdown to avoid hitting the limit. Default 100
requestsPerHour	Maximum requests per hour	int	1000	Maximum count of requests to ChatGPT per hour - from the second half there will be a slowdown to avoid hitting the limit. Default 1000
requestsPerDay	Maximum requests per day	int	3000	Maximum count of requests to ChatGPT per day - from the second half there will be a slowdown to avoid hitting the limit. Default 3000
embeddingsUrl	URL of the embeddings service	String	https://api.openai.com/v1/embeddings	Optional, if not OpenAI's default https://api.openai.com/v1/embeddings
embeddingsModel	Embeddings model	String	text-embedding-3-small	Optional model to use for the embeddings. The default is text-embedding-3-small.

GPT Dictation Service Configuration (base)

Configures whether it's enabled (default false), the model and the request counts, and the maximum request size.

id	name	type	default value	description
disabled	Disabled	boolean	false	Whether the service is disabled.
model	Model	String	whisper-1	The model to use for dictation, default whisper-1.
requestsPerMinute	Maximum requests per minute	int	30	Maximum count of requests to ChatGPT per minute - from the second half there will be a slowdown to avoid hitting the limit. Default 30.
requestsPerHour	Maximum requests per hour	int	100	Maximum count of requests to ChatGPT per hour - from the second half there will be a slowdown to avoid hitting the limit. Default 100.
requestsPerDay	Maximum requests per day	int	300	Maximum count of requests to ChatGPT per day - from the second half there will be a slowdown to avoid hitting the limit. Default 300.
maxRequestSize	Maximum request size in bytes	int	5000000	Maximum request size in bytes, default 5000000. (about one minute of stereo audio with 44.1 kHz and 16 bit)

Composum AI Permission Configuration (slingbase)

A configuration for allowed AI services. There can be multiple configurations, and the allowed services are aggregated. There is a fallback configuration that is used if no other configuration is found, and a factory for multiple configurations which override the fallback configuration if present. If configured, Sling Context Aware Configuration takes precedence over OSGI configuration.

id	name	type	default value	description
10	Services	String[]	[CATEGORIZE, CREATE, SIDEPANEL, TRANSLATE]	List of services to which this configuration applies. Possible values are: CATEGORIZE, CREATE, SIDEPANEL, TRANSLATE. For AEM only create and sidepanel are supported.
20	Allowed Users	String[]	[.*]	Regular expressions for allowed users or user groups. If not present, no user is allowed from this configuration.
30	Denied Users	String[]		Regular expressions for denied users or user groups. Takes precedence over allowed users.
40	Allowed Paths	String[]	[/content/.*]	Regular expressions for allowed content paths. If not present, no paths are allowed.
50	Denied Paths	String[]	[/content/dam/.*]	Regular expressions for denied content paths. Takes precedence over allowed paths.
60	Allowed Views	String[]	[.*]	Regular expressions for allowed views - that is, for URLs like /editor.html/.. If not present, no views are allowed. Use . to allow all views.
70	Denied Views	String[]		Regular expressions for denied views. Takes precedence over allowed views.
80	Allowed Components	String[]	[.*]	Regular expressions for allowed resource types of components. If not present, no components are allowed.
90	Denied Components	String[]		Regular expressions for denied resource types of components. Takes precedence over allowed components.
100	Allowed Page Templates	String[]	[.*]	Regular expressions for allowed page templates. If not present, all page templates are allowed.
110	Denied Page Templates	String[]		Regular expressions for denied page templates. Takes precedence over allowed page templates.

Composum AI Prompt Library Configuration (backend/slingbase)

Location for the prompt library for Composum AI. There can be multiple configurations, and the allowed services are aggregated. There is a fallback configuration that is used if no other configuration is found, and a factory for multiple configurations which override the fallback configuration if present. If configured, Sling Context Aware Configuration takes precedence over OSGI configuration.

id	name	type	default value	description
contentCreationPromptsPath	Content Creation Prompts Path	String		Path to the content creation prompts. Either a JSON file, or a page.
sidePanelPromptsPath	Side Panel Prompts Path	String		Path to the side panel prompts. Either a JSON file, or a page.

Composum AI Translation Service Configuration (backend/base)

Configuration for the basic Composum AI Translation Service

id	name	type	default value	description
disabled	Disable the translation service	boolean	false
fakeTranslation	Fake translation	boolean	false	For quick and inexpensive testing, when you just want to check that the translation does something for e.g. a bulk of texts, you can enable this. The “translation” then just turns the text iNtO tHiS cApItAlIsAtIoN. Easy to spot, but probably doesn't destroy the content completely.
diskCache	Disk cache	String		Path to a directory where to cache the translations. If empty, no caching is done. If the path is relative, it is relative to the current working directory. If the path is absolute, it is used as is.
temperature	temperature	String		The sampling temperature, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
seed	seed	String		If specified, OpenAI will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result.

Composum AI Tool Get Page Markdown (backend/slingbase)

Provides the AI with a tool to search for page paths. Needs a lucene index for all pages. If there is no configuration, the tool is not active.

id	name	type	default value	description
allowedPathsRegex	Allowed paths regex	String	/content/.*	A regex to match the paths that this tool is allowed to be used on. Default: /content/.*

Composum AI Html To Approximate Markdown Service Plugin (slingbase)

A plugin for the ApproximateMarkdownService that transforms the rendered HTML of components to markdown, which can work better than trying to guess the text content from the JCR representation (as is the default) but probably doesn't work for all components. So it can be enabled for some sling resource types by regex. We will not use this for the first two levels below the page, as that could include unwanted stuff like headers and footers.

id	name	type	default value	description
allowedResourceTypes	Allowed resource types	String[]	{“.*”}	Regular expressions for allowed resource types. If not present, no resource types are allowed.
deniedResourceTypes	Denied resource types	String[]	{}	Regular expressions for denied resource types. Takes precedence over allowed resource types.

Composum AI Approximate Markdown Cache Service Configuration (slingbase)

If configured, caches the calculated approximate markdown of pages.
CAUTION: the page content must be independent of the user, or you might leak one user's data to another!

id	name	type	default value	description
disabled	Disable	boolean	false	Disable the service
cacheRootPath	Cache Root Path	String		The JCR root path where the markdown is stored. If not set, no caching is done. Suggestion: /var/composum/ai-markdown-cache. To set this up you'll need to create this path in the repository, add a service user for this bundle's name (composum-ai-integration-backend-slingbase) and make the path writeable for this user.

Composum AI Tool Search Pages (backend/slingbase)

Provides the AI with a tool to search for page paths. Needs a lucene index for all pages. If there is no configuration the tool is not active.

id	name	type	default value	description
resultCount	Result count	int	20	The number of results to return. Default is 20.
siteLevel	Site level	int	2	The number of path segments a site has, used to identify the site root. Default is 2, for sites like /content/my-site.

Composum AI SlingCaConfig Plugin (slingbase)

Allows enabling / disabling the Sling Context Aware Configuration of the Composum AI.

id	name	type	default value	description
1	Enabled	boolean	true	Whether the Sling Context Aware Configuration of the Composum AI is enabled.