Graph Traversal

Graph traversal endpoints let you explore the knowledge graph’s entity-relationship network. You can traverse from a starting entity, search for entities by name, and resolve graph-connected entities to actual searchable documents.

Available Endpoints

Endpoint	Method	Description
/data/v1/query/graph	POST	Traverse graph from an entity
/data/v1/query/graph/search	POST	Search entities by name
/data/v1/query/graph/entity/{name}	GET	Get single entity by exact name
/data/v1/query/graph/documents	POST	Traverse + resolve entities to documents

---

Graph Traversal

Traverse connected entities and relationships from a starting entity.

POST /data/v1/query/graph

Request Parameters

Parameter	Type	Required	Default	Description
entity_name	string	✅ Yes	—	Starting entity name
relationship_types	string[]	No	all	Filter by relationship types
direction	string	No	both	Traversal direction: in, out, both
depth	integer	No	2	How many hops (1–5)
limit	integer	No	50	Maximum entities (1–200)

Example

curl -X POST "https://api.solidrust.ai/data/v1/query/graph" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "entity_name": "Nightmare",
    "relationship_types": ["requires", "has_module"],
    "direction": "both",
    "depth": 2,
    "limit": 50
  }'

import requests

def graph_traversal(entity_name: str, depth: int = 2, rel_types: list[str] = None) -> dict:
    response = requests.post(
        f"{BASE_URL}/data/v1/query/graph",
        headers={"Authorization": f"Bearer {API_KEY}", "Content-Type": "application/json"},
        json={
            "entity_name": entity_name,
            "relationship_types": rel_types,
            "depth": depth,
            "limit": 50,
        },
    )
    return response.json()

result = graph_traversal("Nightmare", depth=2)

print(f"Found {result['total_entities']} entities, {result['total_relationships']} relationships")
for entity in result["entities_found"]:
    print(f"  {entity['name']} ({entity['entity_type']})")
for rel in result["relationships"]:
    print(f"  {rel['source_name']} —[{rel['relationship_type']}]→ {rel['target_name']}")

Response

{
  "entity_name": "Nightmare",
  "entities_found": [
    {"id": "node-1", "name": "Nightmare", "entity_type": "ship", "properties": {}},
    {"id": "node-2", "name": "Caldari Navy", "entity_type": "faction", "properties": {}}
  ],
  "relationships": [
    {"source_name": "Nightmare", "target_name": "Caldari Navy", "relationship_type": "requires", "properties": {}}
  ],
  "total_entities": 2,
  "total_relationships": 1
}

---

Entity Search

Search for entities by name (case-insensitive contains match).

POST /data/v1/query/graph/search

Request Parameters

Parameter	Type	Required	Default	Description
query	string	✅ Yes	—	Search text (substring match)
entity_type	string	No	—	Filter by entity type
limit	integer	No	20	Maximum results (1–100)

Example

def search_entities(query: str, entity_type: str = None) -> dict:
    response = requests.post(
        f"{BASE_URL}/data/v1/query/graph/search",
        headers={"Authorization": f"Bearer {API_KEY}", "Content-Type": "application/json"},
        json={"query": query, "entity_type": entity_type, "limit": 20},
    )
    return response.json()

result = search_entities("night", entity_type="ship")
for entity in result["results"]:
    print(f"  {entity['name']} ({entity['entity_type']})")

---

Get Entity by Name

Look up a single entity by exact name match.

GET /data/v1/query/graph/entity/{entity_name}

curl "https://api.solidrust.ai/data/v1/query/graph/entity/Nightmare" \
  -H "Authorization: Bearer YOUR_API_KEY"

Returns null if the entity is not found.

---

Graph Documents (Traversal + Resolution)

The most powerful graph endpoint — traverses from an entity and resolves connected entities to Milvus documents so you can search them.

POST /data/v1/query/graph/documents

Request Parameters

Parameter	Type	Required	Default	Description
entity_name	string	✅ Yes	—	Starting entity name
entity_type	string	No	—	Optional entity type filter
relationship_types	string[]	No	all	Filter by relationship types
max_depth	integer	No	2	Max traversal depth (1–5)
limit	integer	No	20	Max documents (1–100)
resolve_documents	boolean	No	true	Resolve entities to Milvus documents

Example

def graph_documents(entity_name: str, max_depth: int = 2, limit: int = 20) -> dict:
    response = requests.post(
        f"{BASE_URL}/data/v1/query/graph/documents",
        headers={"Authorization": f"Bearer {API_KEY}", "Content-Type": "application/json"},
        json={
            "entity_name": entity_name,
            "max_depth": max_depth,
            "limit": limit,
            "resolve_documents": True,
        },
    )
    return response.json()

result = graph_documents("Nightmare", max_depth=2)

print(f"Found {result['total']} documents related to '{result['entity_name']}'")
print(f"Related entities: {', '.join(result['related_entities'])}")

for doc in result["results"]:
    print(f"\n{doc['title']} (score: {doc['score']:.2f}, via: {doc['relationship']})")
    print(f"  {doc['content'][:150]}...")

Response

{
  "entity_name": "Nightmare",
  "entity_type": null,
  "results": [
    {
      "id": "doc-abc",
      "title": "Nightmare",
      "content": "The Nightmare is a Caldari Navy battleship...",
      "source": "eve_sde",
      "source_id": "ship_nightmare",
      "content_type": "game_item",
      "url": null,
      "score": 1.0,
      "relationship": "self",
      "related_entity": null
    }
  ],
  "related_entities": ["Caldari Navy", "Shield Extender II"],
  "total": 1
}

Error Responses

Status	Code	Description
400	invalid_request	Missing or invalid parameters
401	INVALID_API_KEY	Missing or invalid API key
503	service_unavailable	Neo4j graph store unavailable

Related

• Knowledge Graph — Legacy entity relationship queries
• Hybrid Search — Automatic graph + vector combination
• Semantic Search — Vector-only search