Knowledge Graph

Track entities, relationships, and connections discovered across browsing sessions

The knowledge graph is an in-memory entity store for tracking companies, people, technologies, and other entities discovered while browsing. Add entities, define relationships between them, merge duplicates, and visualize the graph as a Mermaid diagram. Use entity_query for natural language questions about your collected knowledge.

Tools

Tool	Description	Key Parameters
`entity_query`	Query the graph with natural language	`question` (required)
`entity_add`	Add an entity to the graph	`name` (required), `entity_type` (required), `attributes`
`entity_relate`	Add a relationship between two entities	`from` (required), `to` (required), `relationship` (required)
`entity_merge`	Merge two entities (second into first)	`name_a` (required), `name_b` (required)
`entity_find_related`	Find entities connected to a given entity	`name` (required)
`entity_search`	Fuzzy search entities by name	`query` (required)
`entity_visualize`	Generate a Mermaid diagram of the graph	(none)

Examples

Build a knowledge graph from browsing

{
  "tool": "entity_add",
  "arguments": {
    "name": "Stripe",
    "entity_type": "company",
    "attributes": {
      "industry": "fintech",
      "headquarters": "San Francisco",
      "hiring": true
    }
  }
}

{
  "tool": "entity_relate",
  "arguments": {
    "from": "Stripe",
    "to": "Rust",
    "relationship": "uses"
  }
}

Supported entity types: "company", "person", "technology", "product", "location", "other".

Query with natural language

{
  "tool": "entity_query",
  "arguments": {
    "question": "what companies use Rust?"
  }
}

Returns matching entities and their relationships. The graph persists across the browsing session, accumulating knowledge as you visit more pages.

Visualize the graph

{
  "tool": "entity_visualize",
  "arguments": {}
}

Returns Mermaid graph syntax that renders as an interactive diagram showing all entities and their relationships.