Tools

list_tenants

The accounts you belong to in this region. Claude calls it first.

get_guide

Loads one of the operating guides before Claude works in an area.

get_tenant, get_tenant_context, create_tenant, edit_tenant

Read an account and your role in it, create an account, rename it or set its logo URL.

list_projects, get_project, get_project_context, create_project, edit_project

List the projects you can access, read one with all its assets and your role, create and edit projects — including the terms and conditions app users must accept before they talk to its agents.

list_project_users

The end users who had conversations in a project.

list_tenant_members, set_tenant_member_role, remove_tenant_member †, list_project_members, set_project_member_role, remove_project_member †

The team of an account or project and their roles.

list_tenant_invitations, list_project_invitations, invite_members_to_tenant, invite_members_to_project, revoke_tenant_invitation †, revoke_project_invitation †

Invitations: each invited address receives an email with a link to accept.

list_agents, get_agent, get_agent_revisions

Find agents, read one revision in full, see all revisions with their states.

create_agent, edit_agent

Create an agent as a draft (optionally from a template) and change it; every edit saves a new draft revision.

promote_agent_to_preview, promote_agent_to_published

Move a draft to preview, and a preview to published — live for end users.

revert_agent †, archive_agent †

Go back to an earlier revision; take a revision out of use.

get_voices, get_visuals, list_templates

The voices, avatars and environments an agent can use, and the published agent templates.

list_evaluations, get_evaluation, create_evaluation, edit_evaluation, delete_evaluation †

Evaluations of a project. Changes are saved to the evaluation’s draft revision.

get_evaluation_revisions, promote_evaluation_to_preview, promote_evaluation_to_published, revert_evaluation †

The revisions of an evaluation and their lifecycle: agents run only the published revision, so a change takes effect once it is promoted to preview and then published.

add_insight, edit_insight, remove_insight †

The output data elements of an evaluation: yes/no, a number on a scale, text, or a choice from a list.

add_group, edit_group, remove_group †

Groups of elements that are scored together.

list_knowledge, get_knowledge_spec, get_knowledge

Knowledge sources, their settings, and the text extracted from them.

add_web_knowledge, refresh_web_knowledge

Add a public web page as a knowledge source, or fetch it again.

upload_knowledge_file

Add a document (or replace one) through a one-time upload link (Files, Links and Conversation Data).

edit_knowledge, delete_knowledge †

Change a source’s name, description or availability, or remove it.

list_multimodal_interactions, get_multimodal_interaction

The interactive content of a project: web pages, images, videos, input fields, choices and handovers to other agents.

create_multimodal_interaction, edit_multimodal_interaction, delete_multimodal_interaction †

Create, change and remove elements.

upload_interactive_content_file

Attach the image or video file of an element through a one-time upload link.

list_integrations

The integrations of a project: webhooks before and after a conversation, MCP servers, connectors.

create_pre_integration, create_post_integration, create_mcp_integration, edit_integration, delete_integration †

Manage webhooks and your own MCP servers.

load_mcp_tools

Check an MCP server and list the tools it offers, before adding it.

list_connectors, enable_connector, update_connector, delete_connector †

The connectors Promethist offers, and which of their tools a project uses.

list_project_skills

The skills assigned to a project, which its agents can use.

get_analytics_summary

An overview of a project’s conversations, usage, users, interactive content and evaluations.

get_conversation_analytics, get_usage_analytics, get_user_analytics, get_multimodal_analytics, get_evaluation_analytics, list_analytics_users

The individual analytics, as in the Studio’s analytics page.

export_analytics

The per-session report of a time range — as JSON, or as an Excel or CSV download link.

get_subscription, get_subscription_by_project, get_subscription_by_agent, list_invoices, get_usage

Subscription, invoices and billed usage, read only.

list_sessions, get_session_transcript

Find conversations of a project and read one in full.

search_turns, search_llm_logs, describe_conversation_data

Search the raw record of conversations — individual turns and the model calls behind them — and see which fields it contains.

export_conversation_data

Export raw records as a download link, for datasets too large to page through; the records can be narrowed to named fields.

test_agent

Hold a live text conversation with an agent, one turn per call, to see how it really behaves.

list_test_suites, get_test_suite, create_test_suite, edit_test_suite, delete_test_suite †

Test suites: scenarios that pair an agent with a simulated user.

run_test_suite, get_test_suite_runs, get_test_suite_run

Start a suite and read its conversations, verdicts, scores and transcripts.

save_memory_snapshot, restore_memory_snapshot †

Copy what a simulated user remembers, and put that copy back, so one session of a multi-session test can be re-run without replaying the sessions before it.