Skip to content
Ishvana Ishvana Docs

Conduct Research

POST
/api/agents/pipeline/research/conduct

Conduct comprehensive multi-source research via Lagan.

Uses Lagan’s knowledge processor to perform research and returns results in the same format as Hawken’s research pipeline for frontend compatibility.

Request Body required

Section titled “Request Body required ”
ConductResearchRequest

Request for comprehensive research.

object
message
required
Message

Research query or question

string
>= 1 characters
taskType
Any of:
string
includeLore
Includelore

Include project lore

boolean
default: true
includeSpecialists
Includespecialists

Include specialist help

boolean
default: true

Responses

Section titled “ Responses ”

200

Section titled “200 ”

Successful Response

ResearchPipelineResponse

Response matching the frontend’s ResearchPipelineResult type.

object
researchTerms
Researchterms
Array<string>
ragResults
Ragresults
Array<object>
RagResultItem

Single RAG result.

object
content
required
Content

Document content

string
relevance
required
Relevance

Relevance score

number
<= 1
source
Source

Source name

string
default: knowledge_base
metadata
Metadata
object
key
additional properties
string
projectLore
Projectlore
Array<string>
specialistKnowledge
Specialistknowledge
Array<string>
externalSources
Externalsources
Array<string>
contextSummary
Contextsummary
string
""
relevanceScore
Relevancescore
integer
0 <= 100

422

Section titled “422 ”

Validation Error

HTTPValidationError
object
detail
Detail
Array<object>
ValidationError
object
loc
required
Location
Array
msg
required
Message
string
type
required
Error Type
string
input
Input
ctx
Context
object