Business & Operations

Klarity

by Klarity.AI

Overview

Bring your organization's processes and operational knowledge into your AI assistant. The Klarity MCP connects to your Klarity workspace so you can query your organization's processes, explore the process index knowledge graph linking processes to systems and teams, and ground answers in how your business actually runs — not generic guesses.

Tools

fetch

ChatGPT

Fetch full details for one process: hierarchy placement, current version, steps, policies, linked artifacts, and dependencies.

get_agent_builder_instructions

ChatGPT

Use when the user wants to build a skill, agent, or automation with Klarity, or wants to explore, assess, or audit their organization's processes to find what is worth automating or improving. Also fires on vague asks like 'what can AI do for my team' when a Klarity workspace is connected - the instructions route the conversation from there.

get_agent_builder_resource

ChatGPT

Fetch a Klarity Agent Builder resource (reference, primitive, or template) by type and subtype. Returns the raw content payload and its version.

get_artifact_content

ChatGPT

Retrieve artifact text content. mode='preview' (default) returns capped text up to ~15k tokens. mode='range' returns numbered lines with optional start_line/end_line slice (~5k tokens).

get_artifact_details

ChatGPT

Retrieve artifact metadata and relationships.

get_hierarchy_node_details

ChatGPT

Inspect a process index item including parent, children, and linked process.

get_observation_citation

ChatGPT

Fetch the source evidence and session activity for a process observation. Returns the activity_timeline from the linked recording session, projected to a verbosity tier and optionally narrowed to a single block.

get_process_details

ChatGPT

Fetch process details. Default returns a summary with operational attributes. Pass scope options for heavier sections: - tasks: task/subtask/step tree - observations: recent executions, summary shape - dependencies: downstream processes - linked_artifacts: input/output/related artifact links - policies: policy text - history: prior version summaries

get_process_hierarchy

ChatGPT

Return the process hierarchy tree. Each node includes name, type, and children. Use max_depth to limit tree depth. Omit root to start from top-level.

get_process_observations

ChatGPT

List recorded executions of a process. Each observation surfaces friction signals — where people struggled, what worked, why it was captured.

get_recent_process_changes

ChatGPT

List recent process version changes across the workspace.

list_accessible_workspaces

ChatGPT

List workspaces the authenticated user can access and show which is currently selected.

list_sessions

ChatGPT

List recent workspace sessions with the processes they updated, duration, and the user who created each session.

search

ChatGPT

Search your organization's process index by semantic similarity. Run multiple searches with refined queries — single queries rarely cover broad topics. Use fetch for full process details; use get_process_hierarchy to browse the index.

search_artifact_text

ChatGPT

Search within one document by resource_key, returning surrounding-context matches. For multi-document retrieval, use search_artifacts.

search_artifacts

ChatGPT

Search artifacts using hybrid semantic + lexical retrieval. Returns ranked results with text snippets. Use get_artifact_content for the full document.