The Context Graph is the layer that sits beneath the Process Index and Process observations. It holds the entities, communities, relationships, claims, and text chunks derived from the underlying artifacts — the raw material from which the Process Index is constructed. Querying the graph is querying that raw layer directly.Documentation Index
Fetch the complete documentation index at: https://developers.klarity.ai/llms.txt
Use this file to discover all available pages before exploring further.
What lives in the graph
| Object | What it represents |
|---|---|
| Entity | A specific thing in the organization: a system, team, person, control, policy, or named concept |
| Relationship | A typed edge between two entities — handoff, dependency, ownership, supports, etc. |
| Community | A cluster of entities that frequently appear together; represents a sub-area of how the org works |
| Claim | A specific statement extracted from an artifact, attached to one or more entities |
| Text chunk | The source passage of text behind a claim |
What you ask of it
| User question | Tool to start with |
|---|---|
| ”Find the entity for X” | search_knowledge_graph |
| ”What does System X cover?” | get_entity_details |
| ”What feeds this?” | get_upstream_sources |
| ”What breaks if this changes?” | get_downstream_dependencies |
| ”Map the area around this entity” | explore_graph_neighbors |
| ”Summarize this community” | summarize_community_subgraph |
Always follow up search with detail
search_knowledge_graph returns starting points — entities, communities, claims, chunks — with short snippets. Never answer from those snippets alone. Always chain into one of the get_* or explore_* tools to read the full entity, walk the neighborhood, or pull the supporting text chunk. The snippets are the lookup, not the answer.
See Tools by purpose → Context Graph for the full toolset.