Python SDK and CLI

Revornix provides an official revornix Python package with two layers:

Python SDK: for backend scripts, automations, scheduled jobs, and CI flows
CLI: for shell scripts, terminal usage, and toolchain integration

Source and release info: Revornix-Python-Lib

If you want to integrate Revornix without hand-writing HTTP requests, start here. The package still calls the public Third-Party API underneath.

Installation


pip install revornix

The package currently requires Python >=3.10.

Authentication

Both SDK and CLI require:

base_url: your Revornix API base URL
api_key: your Revornix API key

In Revornix, click the avatar in the bottom-left, then go to Account -> APIKey Management to create or copy a key.

The CLI supports these environment variables:


export REVORNIX_BASE_URL="https://api.revornix.cn"
export REVORNIX_API_KEY="YOUR_API_KEY"

CLI Quick Start

Show help:


revornix --help

If you do not want to rely on environment variables, pass credentials directly:


revornix --base-url "https://api.revornix.cn" --api-key "YOUR_API_KEY" --help

The CLI is organized into 4 command groups:

files: upload files
documents: create, update, search, and delete documents
labels: manage document labels and section labels
sections: manage sections, publish status, and section documents

Upload a file:


revornix files upload \
  --local-file-path ./demo.txt \
  --remote-file-path uploads/demo.txt \
  --content-type text/plain

Create a quick note:


revornix documents create-quick-note \
  --content "hello world" \
  --title "Quick Note" \
  --section 1 \
  --label 10 \
  --auto-summary

Upload a local file and create a file document in one step:


revornix documents upload-create-file \
  --local-file-path ./demo.pdf \
  --remote-file-path uploads/demo.pdf \
  --title "Demo File" \
  --section 1 \
  --label 10

Create a website document:


revornix documents create-website \
  --url https://example.com \
  --section 1 \
  --label 10

Create a section and auto-publish it:


revornix sections create \
  --title "AI Notes" \
  --description "Knowledge base for AI" \
  --process-task-trigger-type 1 \
  --label 10 \
  --auto-publish

CLI Notes

Repeat --section for multiple section IDs
Repeat --label for multiple label IDs
Repeat --document-id when deleting multiple documents
Some booleans are simple flags such as --auto-summary, while others take explicit values such as --status true
CLI responses are printed as JSON, which makes them easy to pipe into scripts
documents search-vector depends on the server already having embeddings for target documents

Python SDK Quick Start

Create a Session:


from revornix import Session
 
session = Session(
    base_url="https://api.revornix.cn",
    api_key="YOUR_API_KEY",
)

Import schema models:


from revornix.schema import DocumentSchema, SectionSchema

Create a quick note document:


from revornix import Session
from revornix.schema import DocumentSchema
 
session = Session(
    base_url="https://api.revornix.cn",
    api_key="YOUR_API_KEY",
)
 
data = DocumentSchema.QuickNoteDocumentParameters(
    title="Quick Note",
    description="Created from SDK",
    cover=None,
    sections=[1],
    labels=[10],
    content="hello world",
    auto_summary=False,
    auto_podcast=False,
    auto_tag=False,
)
 
res = session.create_quick_note_document(data=data)

Upload a file and create a file document:


from revornix import Session
from revornix.schema import DocumentSchema
 
session = Session(
    base_url="https://api.revornix.cn",
    api_key="YOUR_API_KEY",
)
 
session.upload_file(
    local_file_path="./demo.pdf",
    remote_file_path="uploads/demo.pdf",
)
 
data = DocumentSchema.FileDocumentParameters(
    title="Demo File",
    description="Imported from uploaded file",
    cover=None,
    sections=[1],
    labels=[10],
    file_name="uploads/demo.pdf",
    auto_summary=False,
    auto_podcast=False,
    auto_tag=False,
)
 
res = session.create_file_document(data=data)

Run vector search:


from revornix import Session
from revornix.schema import DocumentSchema
 
session = Session(
    base_url="https://api.revornix.cn",
    api_key="YOUR_API_KEY",
)
 
data = DocumentSchema.VectorSearchRequest(query="retrieval augmented generation")
res = session.search_document_vector(data=data)

Coverage

Session already covers the main operations:

file upload
document create/detail/update/delete/search/vector-search
document label and section label create/list/delete
section create/detail/update/delete/search/publish/republish

If you need lower-level field-by-field definitions, continue with Third-Party API.

Good Fits

bulk ingestion into Revornix from Python scripts
backend jobs, CI, or cron-based content workflows
a lower-level integration layer for OpenClaw, workflow engines, or custom agents