Integration

AITEZA integration

AITEZA is a German AI platform for company documents: datarooms per customer or project, answers with a source for every statement. Create and manage datarooms, upload files, images and websites into them (text is extracted and embedded), search their content, ask questions in chats grounded in datarooms and files, translate documents, manage versions, members, groups and models, keep a prompt and skill library, and register QM or n8n workflows. Not exposed: administration, Relay guest sharing, API key management, cloud-storage sync, MCP integrations, personal preferences, ZIP downloads, skill import and export, streaming answers, and the delete-all routes for a dataroom's files, images or websites. Authentication: a personal API key (starts with azk_) created in AITEZA under Settings, API keys (needs the API keys role), sent in the X-API-Key header, plus the address of your AITEZA installation (the AITEZA cloud or your own server). A key acts as you in one organization without your admin rights; a Read key only reads, and keys cannot run workflows. At most 60 requests per minute per key unless the installation sets another limit.

Connect AITEZA How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in AITEZA.Built from the 151 AITEZA tools below, on the connection you add once.

What you can do with AITEZA

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Each badge shows the default agent permission of an endpoint (allowed: an agent calls it without asking; ask first: you are asked before an agent calls it, when the agent or chat asks permission for sensitive actions; never: no agent may call it), which every user can change endpoint by endpoint in their credential settings.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

File, 15 operations

assign_file_to_dataroomAssign file to dataroomPATCHAgents: ask firstMoves one of your standalone files into a dataroom (editor access needed); embeddings are created in the background (processed stays false until done).
assign_generated_file_to_dataroomAssign generated file to dataroomPATCHAgents: ask firstMoves a file AITEZA generated (for example in a chat) into a dataroom without copying or re-extracting it; a PDF preview and embeddings are made in the background.
delete_dataroom_fileDelete dataroom fileDELETEAgents: ask firstDeletes one file of a dataroom with its embeddings (cannot be undone).
delete_dataroom_filesDelete dataroom filesDELETEAgents: ask firstDeletes several files of a dataroom by id, with their embeddings (cannot be undone).
delete_fileDelete fileDELETEAgents: ask firstDeletes one of your standalone files permanently, from storage and database (cannot be undone).
get_dataroom_fileGet dataroom fileGETAgents: ask firstReturns one file of a dataroom with its processed Markdown content (and edited content), status, failure reason, malware scan result and audio transcript.
get_fileGet fileGETAgents: ask firstReturns one of your standalone files with its processed content, status and failure reason.
get_file_contentGet file contentGETAgents: ask firstReturns the processed text content of one of your standalone files.
get_file_statusGet file statusGETAgents: ask firstReturns the processing status of one of your standalone files (status, processed, failed); poll it after upload_file until processed.
list_dataroom_filesList dataroom filesGETAgents: ask firstLists the files of a dataroom (without content), 0-based pages of up to 200, with processing status; filter by name or uploader (q), processed, or AI-generated.
list_filesList filesGETAgents: ask firstLists your standalone files (uploaded or AI-generated, not in a dataroom), 0-based pages of up to 200, optionally filtered by name.
move_dataroom_fileMove dataroom filePATCHAgents: ask firstMoves one file to another dataroom; needs permissions on both datarooms.
move_dataroom_filesMove dataroom filesPUTAgents: ask firstMoves several files from a dataroom to another dataroom.
reprocess_dataroom_fileReprocess dataroom filePOSTAgents: ask firstReprocesses one file with a pipeline (for example OCR for a scan) to rebuild its content and embeddings, or starts processing of a file stored unprocessed.
reprocess_dataroom_filesReprocess dataroom filesPATCHAgents: ask firstReprocesses several files of a dataroom with a pipeline (reprocess true). The whole request is refused (415) when a selected file has a type AITEZA cannot extract text from.

Image, 12 operations

assign_image_to_dataroomAssign image to dataroomPATCHAgents: ask firstMoves one of your standalone images into a dataroom (editor access needed); descriptions are generated if missing, then embeddings.
delete_dataroom_imageDelete dataroom imageDELETEAgents: ask firstDeletes one image of a dataroom with its embeddings (cannot be undone).
delete_dataroom_imagesDelete dataroom imagesDELETEAgents: ask firstDeletes several images of a dataroom by id, with their embeddings (cannot be undone).
delete_imageDelete imageDELETEAgents: ask firstDeletes one of your standalone images permanently (cannot be undone).
get_dataroom_imageGet dataroom imageGETAgents: ask firstReturns one image of a dataroom: size, dimensions, short and detailed AI descriptions, source file and page, processing status.
get_imageGet imageGETAgents: ask firstReturns one of your standalone images with its AI descriptions and processing status.
get_image_statusGet image statusGETAgents: ask firstReturns the processing status of one of your standalone images; poll it after upload_image until processed.
list_dataroom_imagesList dataroom imagesGETAgents: ask firstLists the images of a dataroom with their AI descriptions and processing status, 0-based pages; filter by name or uploader (q) or processed.
list_imagesList imagesGETAgents: ask firstLists your standalone images (not in a dataroom), 0-based pages of up to 200, optionally filtered by name.
move_dataroom_imageMove dataroom imagePATCHAgents: ask firstMoves one image to another dataroom; needs permissions on both datarooms.
move_dataroom_imagesMove dataroom imagesPUTAgents: ask firstMoves several images from a dataroom to another dataroom.
reprocess_dataroom_imagesReprocess dataroom imagesPATCHAgents: ask firstRegenerates the descriptions and embeddings of several images (reprocess true), or renames them all with name.

Dataroom, 9 operations

copy_dataroomCopy dataroomPOSTAgents: ask firstCopies a dataroom and the subdatarooms you can view, with their finished files, websites and images, under a new name and parent (all for the root level). You own the copy.
create_dataroomCreate dataroomPOSTAgents: ask firstCreates a dataroom (optionally under a parent) with access private, inherit (from the parent) or custom (then list owners, editors, viewers and groups), and returns it.
delete_dataroomsDelete dataroomsDELETEAgents: ask firstDeletes one or more datarooms with all their files, images and web sources (cannot be undone).
get_connected_dataroomsGet connected dataroomsGETAgents: ask firstReturns the datarooms connected by default to a dataroom's chats: all configured ids, and the ones you may open.
get_dataroomGet dataroomGETAgents: ask firstReturns one dataroom: name, description, custom instructions, models, owners, editors and viewers, tags, parent, counts of files, websites, images and children.
list_child_dataroomsList child dataroomsGETAgents: ask firstLists the child datarooms of a dataroom; pass all as dataroom for the root-level datarooms. Optional filters: name text, empty, owned.
list_recent_dataroomsList recent dataroomsGETAgents: ask firstLists the datarooms you used most recently.
list_starred_dataroomsList starred dataroomsGETAgents: ask firstLists the datarooms you starred, optionally filtered by name.
move_dataroomsMove dataroomsPATCHAgents: ask firstMoves several datarooms under a new parent dataroom, or to the root level with parent all.

Website, 9 operations

add_dataroom_websiteAdd dataroom websitePOSTAgents: ask firstAdds a web page (optionally with its subpages) to a dataroom and starts crawling it; when many crawls run it waits in a queue. Returns the web source.
delete_dataroom_websiteDelete dataroom websiteDELETEAgents: ask firstDeletes one web source of a dataroom (cannot be undone).
delete_dataroom_websitesDelete dataroom websitesDELETEAgents: ask firstDeletes several web sources of a dataroom by id (cannot be undone).
get_dataroom_websiteGet dataroom websiteGETAgents: ask firstReturns one web source of a dataroom: URL, crawl settings, number of pages, schedule and status.
list_dataroom_websitesList dataroom websitesGETAgents: ask firstLists the web sources of a dataroom with crawl settings, schedule and status, 0-based pages; filter by URL or name (q) or processed.
move_dataroom_websiteMove dataroom websitePATCHAgents: ask firstMoves one web source to another dataroom; needs permissions on both datarooms.
move_dataroom_websitesMove dataroom websitesPUTAgents: ask firstMoves several web sources from a dataroom to another dataroom.
rescan_dataroom_websiteRescan dataroom websitePATCHAgents: ask firstRescans one web source now, optionally changing its URL, name or crawl settings (fields you leave out keep their values).
rescan_dataroom_websitesRescan dataroom websitesPATCHAgents: ask firstRescans several web sources (reprocess true), optionally changing the name or crawl settings of all of them.

Chat, 6 operations

create_chatCreate chatPOSTAgents: ask firstCreates an empty chat conversation with an optional name and returns it with its id; ask in it with generate_chat_answer.
delete_chatDelete chatDELETEAgents: ask firstDeletes a chat and all its messages (cannot be undone).
get_chatGet chatGETAgents: ask firstReturns one chat with all its messages, sources and token usage.
list_chatsList chatsGETAgents: ask firstLists your chat conversations (without messages), newest first, 0-based pages of up to 200; optional q filters by chat name.
list_dataroom_chatsList dataroom chatsGETAgents: ask firstLists the chats in which a dataroom was connected, 0-based pages; optional q filters by chat name.
rename_chatRename chatPATCHAgents: ask firstRenames a chat.

Chat message, 6 operations

ask_without_savingAsk without savingPOSTAgents: ask firstRuns one question through the retrieval pipeline without creating or saving a chat, and returns the answer text. For one-off evaluations or classification. Uses AI tokens.
delete_chat_messagesDelete chat messagesDELETEAgents: ask firstDeletes all messages of a chat but keeps the chat itself (cannot be undone).
generate_chat_answerGenerate chat answerPOSTAgents: ask firstAsks the model in an existing chat (create_chat first) and waits for the answer, in result.output.text; the answer is saved in the chat. Uses AI tokens billed to your organization.
get_chat_messageGet chat messageGETAgents: ask firstReturns one message of a chat with its sources and token usage.
get_latest_chat_messageGet latest chat messageGETAgents: ask firstReturns the most recent message of a chat, usually the last answer, with its sources and token usage.
list_chat_messagesList chat messagesGETAgents: ask firstLists every message of a chat in order: prompts and answers with their sources, model, token counts and cost.

File url, 5 operations

get_dataroom_file_download_urlGet dataroom file download urlGETAgents: ask firstReturns a temporary signed URL to download the original uploaded file. Owners and editors only.
get_dataroom_file_preview_urlGet dataroom file preview urlGETAgents: ask firstReturns a temporary signed URL to a PDF rendition of the file (PDFs directly; Office and OpenDocument files once their rendition exists).
get_file_download_urlGet file download urlGETAgents: ask firstReturns a temporary signed URL to download the original of one of your standalone files.
get_file_preview_urlGet file preview urlGETAgents: ask firstReturns a temporary signed URL to a PDF rendition of one of your standalone files.
get_generated_file_version_download_urlGet generated file version download urlGETAgents: ask firstReturns a temporary signed URL to download one version of a generated file.

File version, 5 operations

get_dataroom_file_versionGet dataroom file versionGETAgents: ask firstReturns one version of a file with its content.
get_generated_file_versionGet generated file versionGETAgents: ask firstReturns the Markdown that one version of a generated file was rendered from.
list_dataroom_file_versionsList dataroom file versionsGETAgents: ask firstLists the version history of a file, newest first, without content, 0-based pages.
list_generated_file_versionsList generated file versionsGETAgents: ask firstLists the versions of a file generated in a chat, newest first; a file never updated has the single version 1.
restore_dataroom_file_versionRestore dataroom file versionPOSTAgents: ask firstRestores a file's content and binary to an earlier version; the current state is kept in the history first.

Model, 5 operations

get_modelGet modelGETAgents: ask firstReturns one AI model with capabilities, hosting and prices per million tokens.
list_chat_modelsList chat modelsGETAgents: ask firstLists the AI models you may use in chats, with capabilities, hosting and prices per million tokens.
list_dataroom_modelsList dataroom modelsGETAgents: ask firstLists the AI models assigned to a dataroom that you may use.
list_modelsList modelsGETAgents: ask firstLists the AI models you may use in your organization, with capabilities, hosting and prices per million tokens; exactly one has defaultModel true (your default).
list_vlm_modelsList vlm modelsGETAgents: ask firstLists the vision models usable as vlmModel with the VLM processing pipelines.

Prompt, 5 operations

create_promptCreate promptPOSTAgents: ask firstCreates a prompt in the library (only message is required); with skillName it becomes a skill. Sharing with people or the whole organization (isPublic) makes it visible to them.
delete_promptDelete promptDELETEAgents: ask firstDeletes one prompt (owner access needed; cannot be undone).
delete_promptsDelete promptsDELETEAgents: ask firstDeletes the listed prompts you own (cannot be undone); prompts you do not own are reported as skipped.
get_promptGet promptGETAgents: ask firstReturns one prompt or skill you can see: text, skill name and files, tags, sharing and your permissions.
list_promptsList promptsGETAgents: ask firstLists prompts and skills you created, that were shared with you or are public in your organization, 0-based pages; search q, filter by tags, bookmarked, owned or public.

Workflow, 5 operations

create_workflowCreate workflowPOSTAgents: ask firstRegisters a workflow in AITEZA: a custom workflow started through the AITEZA Trigger in n8n, or a QM workflow with criteria. API keys cannot run workflows.
delete_workflowsDelete workflowsDELETEAgents: ask firstDeletes the listed workflows you own (cannot be undone); reports how many were deleted and which failed.
get_workflowGet workflowGETAgents: ask firstReturns one workflow: category, instructions, default model and sources, criteria, webhook path and version (never the API key).
list_starred_workflowsList starred workflowsGETAgents: ask firstLists the workflows you starred, 0-based pages.
list_workflowsList workflowsGETAgents: ask firstLists the AITEZA workflows you can see (QM and custom), with filters by name or content, categories, tags, shared and starred, and page info.

Dataroom member, 4 operations

add_dataroom_memberAdd dataroom memberPOSTAgents: ask firstGives people of your organization a role in a dataroom (owner, editor or viewer), granting them access, optionally in every child dataroom too.
leave_dataroomLeave dataroomDELETEAgents: ask firstRemoves yourself from a dataroom; you lose access to it.
list_dataroom_membersList dataroom membersGETAgents: ask firstLists the people with a role in a dataroom: username, name, email (personal data) and role; optional q filters by name.
remove_dataroom_memberRemove dataroom memberDELETEAgents: ask firstRemoves people from a dataroom so they lose access, optionally from its child datarooms too.

Tag, 4 operations

list_dataroom_tagsList dataroom tagsGETAgents: ask firstLists the distinct tags used on datarooms in your organizations (lower-case). The full shared vocabulary is list_tags.
list_prompt_tagsList prompt tagsGETAgents: ask firstLists the distinct tags used on prompts in your organizations (lower-case). The full shared vocabulary is list_tags.
list_tagsList tagsGETAgents: ask firstLists the distinct tags used in your organizations (lower-case), optionally for one kind of object; values can be sent back as filters as they are.
list_workflow_tagsList workflow tagsGETAgents: ask firstLists the tags used on the workflows you can see.

Dataroom group, 3 operations

add_dataroom_groupAdd dataroom groupPOSTAgents: ask firstGives groups of your organization a role in a dataroom, so their members get access; optionally in every child dataroom too.
list_dataroom_groupsList dataroom groupsGETAgents: ask firstLists the groups assigned to a dataroom with their role; optional q filters by group name.
remove_dataroom_groupRemove dataroom groupDELETEAgents: ask firstRemoves groups from every role in a dataroom (their members lose that access), optionally in child datarooms too.

Dataroom change, 2 operations

acknowledge_dataroom_changesAcknowledge dataroom changesPOSTAgents: ask firstMarks every change of a dataroom up to and including change id upTo as seen by you, so list_dataroom_changes with unacknowledgedOnly no longer lists them.
list_dataroom_changesList dataroom changesGETAgents: ask firstLists what changed in a dataroom and its subdatarooms, newest first: sources uploaded, replaced, deleted or moved, by users or Relay guests (guest emails are personal data).

File chunk, 2 operations

list_dataroom_file_chunksList dataroom file chunksGETAgents: ask firstLists the chunks of a file (the passages used for retrieval), 0-based pages.
list_file_chunksList file chunksGETAgents: ask firstLists the chunks of a file (passages used for retrieval), 0-based pages; works for your own files and for files of datarooms you can view.

Image url, 2 operations

get_dataroom_image_download_urlGet dataroom image download urlGETAgents: ask firstReturns a temporary signed URL to download the original uploaded image.
get_image_download_urlGet image download urlGETAgents: ask firstReturns a temporary signed URL to download one of your standalone images.

Image version, 2 operations

get_dataroom_image_versionGet dataroom image versionGETAgents: ask firstReturns one version of an image with its content.
list_dataroom_image_versionsList dataroom image versionsGETAgents: ask firstLists the version history of an image, newest first, without content, 0-based pages.

Prompt file, 2 operations

delete_prompt_fileDelete prompt fileDELETEAgents: ask firstRemoves one file bundled with a skill (editor access needed) and returns the prompt.
get_prompt_fileGet prompt fileGETAgents: ask firstDownloads one file bundled with a skill you can see, as a stored file.

Transcript, 2 operations

correct_transcriptCorrect transcriptPATCHAgents: ask firstReplaces the text of every utterance of a transcribed recording (to fix misheard words), in order; an empty text removes that utterance. Saved as a new version.
name_transcript_speakerName transcript speakerPATCHAgents: ask firstGives an anonymous speaker of a transcribed recording (Speaker 1, ...) a display name; the transcript is re-rendered and re-embedded. An empty name removes it.

Website version, 2 operations

get_dataroom_website_versionGet dataroom website versionGETAgents: ask firstReturns one version of a web source with its crawled content.
list_dataroom_website_versionsList dataroom website versionsGETAgents: ask firstLists the version history of a web source, newest first, without content, 0-based pages.

Admin right, 1 operation

get_my_admin_rightsGet my admin rightsGETAgents: ask firstReturns your admin areas (usage, people, audit, members) in the organization: tenantAdmin, the areas granted organization-wide, and per group. API keys themselves never use admin rights.

Chat cost, 1 operation

estimate_chat_costEstimate chat costPOSTAgents: ask firstEstimates the cost of a chat request before sending it: token counts and costs per request, day and month for the chosen model and context. Returns an empty object when cost data is unavailable.

Chat generation, 1 operation

get_chat_generationGet chat generationGETAgents: ask firstReturns the answer currently being generated in a chat, or one that finished in the last minutes: prompt, text, tool calls, generated files, pending questions and outcome. Empty when nothing is generating.

Chat question, 1 operation

answer_model_questionAnswer model questionPOSTAgents: ask firstAnswers a question the model asked while generating (toolApprovals of get_chat_generation): free text for a QUESTION, ALLOW, ALLOW_ALWAYS or DENY for a WEB_SEARCH permission.

File processing, 1 operation

list_file_processing_attemptsList file processing attemptsGETAgents: ask firstLists the stages of a file's latest processing run, oldest first: pipeline tried, outcome, failure reason and why no further stage ran.

Group, 1 operation

list_my_groupsList my groupsGETAgents: ask firstLists the groups you belong to in the organization, by name and path.

Organization, 1 operation

list_my_organizationsList my organizationsGETAgents: ask firstLists your organizations (tenants) with their booked add-ons; an API key sees only its own organization.

Processing pipeline, 1 operation

list_processing_pipelinesList processing pipelinesGETAgents: ask firstLists the document processing pipelines available on this installation (for example DOCLING, VLM), as names.

Prompt access, 1 operation

get_prompt_accessGet prompt accessGETAgents: ask firstReturns who has access to a prompt (owners, editors, viewers, groups, public flag) and whether you may edit or share it.

Usage, 1 operation

get_my_usageGet my usageGETAgents: ask firstReturns your AI cost, tokens and requests for one UTC month: total, per model, per organization and per day, plus a month-end forecast. available false when cost data is unavailable.

Usage limit, 1 operation

get_usage_limitsGet usage limitsGETAgents: ask firstReturns your monthly limits per person (plan name, resources with limit and used this UTC month, reset date) set by the organization that pays for you.

Website url, 1 operation

list_dataroom_website_urlsList dataroom website urlsGETAgents: ask firstLists the URLs of the subpages crawled for a web source, 0-based pages.

Workflow access, 1 operation

get_workflow_accessGet workflow accessGETAgents: ask firstReturns the people and groups with access to a workflow and their roles (names and emails are personal data).

Showing the first 120 of 151 endpoints. The rest are in the product, searchable from the workflow builder and from chat.

Use AITEZA from Claude Code, Cursor or Codex

Connect AITEZA in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its AITEZA operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect AITEZA

Create your workspace
Sign up free. Nothing to install.
Connect AITEZA once
Add your AITEZA connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.