diff --git a/README.md b/README.md index 2075365..fb02513 100644 --- a/README.md +++ b/README.md @@ -29,6 +29,8 @@ This project spawns an API server that can be integrated to provide custom proce After starting the service you can navigate to `http://{paperless-llm-workflows.ip}:8123/api/` to get an up to date API documentation describing all the endpoints. +If you wish to inspect the documentation online here is a [preview link](https://redocly.github.io/redoc/?url=https://raw.githubusercontent.com/ju6ge/paperless-field-extractor/refs/heads/master/openapi.json). + To integrate a functionality into paperless you need add it as webhook trigger in your paperless worflows: ![Paperless Webhook](./example-workflow-action.png) @@ -38,6 +40,11 @@ When a webhook gets triggered the document will be added to the processing queue be updated with the results and is given a `finished` tag. If you wish to assign a specific tag on process completion there is an extra parameter too the webhook which you can use to overwrite what tag will be assigned once the processing step has completed. +As of now the following llm workflows are availible: +- `/fill/custom_fields`: For all supported custom fields datatypes extract value from document content +- `/suggest/correspondent`: Suggest document correspondent by using reasoning + + # Configuration Configuration of the software is possible via a configuration file at `/etc/paperless-field-extractor/config.toml` or via environment variables. Environment variables can be used to overwrite values from the configuration file.