Skip to main content
The projects endpoints allow the Flutter client to notify the backend when new documents have been generated. Each document is split into chunks and upserted into a per-project ChromaDB collection so it can be retrieved as context in future chat requests.

​
POST /api/v1/projects//documents/ingest

Splits a markdown document into chunks and upserts them into the per-project ChromaDB vector store. This operation is idempotent — calling it again with the same content is safe.

​
Path parameters

​
project_id
string
required
The unique identifier for the project. This value scopes the vector store collection so documents from different projects do not mix.

​
Request body

​
doc_name
string
required
Human-readable document name, for example PROJECT_MANIFESTO.md. Between 1 and 255 characters.
​
markdown_content
string
required
Full markdown text of the document to ingest. Must be at least 1 character.

​
Response

{
  "project_id": "7c9e6679-7425-40de-944b-e07fc1f90ae7",
  "doc_name": "PROJECT_MANIFESTO.md",
  "chunks_ingested": 12
}
​
project_id
string
The target project identifier, echoed from the path parameter.
​
doc_name
string
Name of the ingested document, echoed from the request body.
​
chunks_ingested
integer
Number of text chunks stored in the vector collection. Depends on document length and the chunking configuration.

​
Errors

StatusCondition
400 Bad RequestRequest payload is invalid (e.g., empty markdown_content)
500 Internal Server ErrorBackend ingestion failure — check server logs

​
Example

curl -X POST \
  http://localhost:8000/api/v1/projects/7c9e6679-7425-40de-944b-e07fc1f90ae7/documents/ingest \
  -H "Content-Type: application/json" \
  -d '{
    "doc_name": "PROJECT_MANIFESTO.md",
    "markdown_content": "# Project Manifesto\n\nThis project aims to...\n"
  }'

{
  "project_id": "7c9e6679-7425-40de-944b-e07fc1f90ae7",
  "doc_name": "PROJECT_MANIFESTO.md",
  "chunks_ingested": 3
}

Build docs developers (and LLMs) love

Get started for freeTalk to us