gmakstutis/cognee
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Remove all references to SearchType.INSIGHTS across the codebase, meaningfully replacing it with SearchType.GRAPH_COMPLETION where applicable.

Browse source
This commit is contained in:
Daulet Amirkhanov 2025-10-07 16:09:11 +01:00
parent 63a1463073
commit 97f90c95ed
12 changed files with 3 additions and 53 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
cognee-frontend/src/modules/chat/hooks/useChat.ts
View file

@ -89,15 +89,6 @@ export default function useChat(dataset: Dataset) {
}
interface Node {
name: string;
}
interface Relationship {
relationship_name: string;
}
type InsightMessage = [Node, Relationship, Node];
// eslint-disable-next-line @typescript-eslint/no-explicit-any
function convertToSearchTypeOutput(systemMessage: any[] | any, searchType: string): string {
@ -106,14 +97,6 @@ function convertToSearchTypeOutput(systemMessage: any[] | any, searchType: strin
}
switch (searchType) {
case "INSIGHTS":
return systemMessage.map((message: InsightMessage) => {
const [node1, relationship, node2] = message;
if (node1.name && node2.name) {
return `${node1.name} ${relationship.relationship_name} ${node2.name}.`;
}
return "";
}).join("\n");
case "SUMMARIES":
return systemMessage.map((message: { text: string }) => message.text).join("\n");
case "CHUNKS":

2
cognee-mcp/README.md
View file

@ -266,7 +266,7 @@ The MCP server exposes its functionality through tools. Call them from any MCP c
- **codify**: Analyse a code repository, build a code graph, stores it in memory
- **search**: Query memory – supports GRAPH_COMPLETION, RAG_COMPLETION, CODE, CHUNKS, INSIGHTS
- **search**: Query memory – supports GRAPH_COMPLETION, RAG_COMPLETION, CODE, CHUNKS
- **list_data**: List all datasets and their data items with IDs for deletion operations

11
cognee-mcp/src/server.py
View file

@ -478,11 +478,6 @@ async def search(search_query: str, search_type: str) -> list:
Best for: Direct document retrieval, specific fact-finding.
Returns: LLM responses based on relevant text chunks.
**INSIGHTS**:
Structured entity relationships and semantic connections.
Best for: Understanding concept relationships, knowledge mapping.
Returns: Formatted relationship data and entity connections.
**CHUNKS**:
Raw text segments that match the query semantically.
Best for: Finding specific passages, citations, exact content.
@ -524,7 +519,6 @@ async def search(search_query: str, search_type: str) -> list:
- "RAG_COMPLETION": Returns an LLM response based on the search query and standard RAG data
- "CODE": Returns code-related knowledge in JSON format
- "CHUNKS": Returns raw text chunks from the knowledge graph
- "INSIGHTS": Returns relationships between nodes in readable format
- "SUMMARIES": Returns pre-generated hierarchical summaries
- "CYPHER": Direct graph database queries
- "FEELING_LUCKY": Automatically selects best search type
@ -537,7 +531,6 @@ async def search(search_query: str, search_type: str) -> list:
A list containing a single TextContent object with the search results.
The format of the result depends on the search_type:
- **GRAPH_COMPLETION/RAG_COMPLETION**: Conversational AI response strings
- **INSIGHTS**: Formatted relationship descriptions and entity connections
- **CHUNKS**: Relevant text passages with source metadata
- **SUMMARIES**: Hierarchical summaries from general to specific
- **CODE**: Structured code information with context
@ -547,7 +540,6 @@ async def search(search_query: str, search_type: str) -> list:
Performance & Optimization:
- **GRAPH_COMPLETION**: Slower but most intelligent, uses LLM + graph context
- **RAG_COMPLETION**: Medium speed, uses LLM + document chunks (no graph traversal)
- **INSIGHTS**: Fast, returns structured relationships without LLM processing
- **CHUNKS**: Fastest, pure vector similarity search without LLM
- **SUMMARIES**: Fast, returns pre-computed summaries
- **CODE**: Medium speed, specialized for code understanding
@ -586,9 +578,6 @@ async def search(search_query: str, search_type: str) -> list:
return str(search_results[0])
elif search_type.upper() == "CHUNKS":
return str(search_results)
elif search_type.upper() == "INSIGHTS":
results = retrieved_edges_to_string(search_results)
return results
else:
return str(search_results)

1
cognee/api/v1/responses/default_tools.py
View file

@ -14,7 +14,6 @@ DEFAULT_TOOLS = [
"type": "string",
"description": "Type of search to perform",
"enum": [
"INSIGHTS",
"CODE",
"GRAPH_COMPLETION",
"NATURAL_LANGUAGE",

2
cognee/api/v1/responses/dispatch_function.py
View file

@ -59,7 +59,7 @@ async def handle_search(arguments: Dict[str, Any], user) -> list:
valid_search_types = (
search_tool["parameters"]["properties"]["search_type"]["enum"]
if search_tool
else ["INSIGHTS", "CODE", "GRAPH_COMPLETION", "NATURAL_LANGUAGE"]
else ["CODE", "GRAPH_COMPLETION", "NATURAL_LANGUAGE"]
)
if search_type_str not in valid_search_types:

1
cognee/api/v1/responses/routers/default_tools.py
View file

@ -14,7 +14,6 @@ DEFAULT_TOOLS = [
"type": "string",
"description": "Type of search to perform",
"enum": [
"INSIGHTS",
"CODE",
"GRAPH_COMPLETION",
"NATURAL_LANGUAGE",

9
cognee/api/v1/search/search.py
View file

@ -52,11 +52,6 @@ async def search(
Best for: Direct document retrieval, specific fact-finding.
Returns: LLM responses based on relevant text chunks.
**INSIGHTS**:
Structured entity relationships and semantic connections.
Best for: Understanding concept relationships, knowledge mapping.
Returns: Formatted relationship data and entity connections.
**CHUNKS**:
Raw text segments that match the query semantically.
Best for: Finding specific passages, citations, exact content.
@ -124,9 +119,6 @@ async def search(
**GRAPH_COMPLETION/RAG_COMPLETION**:
[List of conversational AI response strings]
**INSIGHTS**:
[List of formatted relationship descriptions and entity connections]
**CHUNKS**:
[List of relevant text passages with source metadata]
@ -146,7 +138,6 @@ async def search(
Performance & Optimization:
- **GRAPH_COMPLETION**: Slower but most intelligent, uses LLM + graph context
- **RAG_COMPLETION**: Medium speed, uses LLM + document chunks (no graph traversal)
- **INSIGHTS**: Fast, returns structured relationships without LLM processing
- **CHUNKS**: Fastest, pure vector similarity search without LLM
- **SUMMARIES**: Fast, returns pre-computed summaries
- **CODE**: Medium speed, specialized for code understanding

4
cognee/cli/commands/search_command.py
View file

@ -31,10 +31,6 @@ Search Types & Use Cases:
Traditional RAG using document chunks without graph structure.
Best for: Direct document retrieval, specific fact-finding.
**INSIGHTS**:
Structured entity relationships and semantic connections.
Best for: Understanding concept relationships, knowledge mapping.
**CHUNKS**:
Raw text segments that match the query semantically.
Best for: Finding specific passages, citations, exact content.

1
cognee/cli/config.py
View file

@ -19,7 +19,6 @@ COMMAND_DESCRIPTIONS = {
SEARCH_TYPE_CHOICES = [
"GRAPH_COMPLETION",
"RAG_COMPLETION",
"INSIGHTS",
"CHUNKS",
"SUMMARIES",
"CODE",

5
cognee/infrastructure/llm/prompts/search_type_selector_prompt.txt
View file

@ -10,8 +10,6 @@ Here are the available `SearchType` tools and their specific functions:
- Summarizing large amounts of information
- Quick understanding of complex subjects
* **`INSIGHTS`**: The `INSIGHTS` search type discovers connections and relationships between entities in the knowledge graph.
**Best for:**
- Discovering how entities are connected
@ -95,9 +93,6 @@ Here are the available `SearchType` tools and their specific functions:
Query: "Summarize the key findings from these research papers"
Response: `SUMMARIES`
Query: "What is the relationship between the methodologies used in these papers?"
Response: `INSIGHTS`
Query: "When was Einstein born?"
Response: `CHUNKS`

2
cognee/modules/retrieval/utils/description_to_codepart_search.py
View file

@ -62,7 +62,7 @@ async def code_description_to_code_part(
try:
if include_docs:
search_results = await search(query_text=query, query_type="INSIGHTS")
search_results = await search(query_text=query, query_type="GRAPH_COMPLETION")
concatenated_descriptions = " ".join(
obj["description"]

1
cognee/tests/cli_tests/cli_unit_tests/test_cli_utils.py
View file

@ -53,7 +53,6 @@ class TestCliConfig:
expected_types = [
"GRAPH_COMPLETION",
"RAG_COMPLETION",
"INSIGHTS",
"CHUNKS",
"SUMMARIES",
"CODE",