Add support for viewer-based credentials for the Cortex chatbot #150
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This commit drops all of the Snowflake-specific authentication logic in favour of using the new `snowflakeauth` package. This has a big advantage in that it works with Snowflake connections and credentials created by the Snowflake Python packages and their CLI. However, this is a breaking change in that (1) some environment variables that previously furnished credentials no longer do; and (2) the signature of `chat_cortex()` has changed. The reason to drop the environment variable support is that these specific environment variables are not used by Snowflake's Python tooling -- so I think this is the right choice. Existing unit tests do cover regressions in the Authorization headers we pass to Snowflake, though I did have to modify them a bit to account for the API changes. Signed-off-by: Aaron Jacobs <aaron.jacobs@posit.co>
This commit ports `odbc`'s recently-added support for "viewer-based
credentials" on Posit Connect to `elmer`. Specifically, passing a Shiny
session object to `chat_cortex()` now allows Cortex chatbots to use the
*viewer's* Snowflake credentials on Connect rather than requiring a
service account or the publisher to embed their personal credentials.
For example:
library(shiny)
library(shinychat)
library(elmer)
stage <- "CORTEX_ANALYST_DEMO.REVENUE_TIMESERIES.RAW_DATA"
model_file <- "revenue_timeseries.yaml"
ui <- bslib::page_fluid(chat_ui("chat", "What questions can I ask?"))
server <- function(input, output, session) {
chat <- chat_cortex(
model_file = glue::glue("@{stage}/{model_file}"),
account = "myaccount",
session = session
)
chat_append("chat", glue::glue(
"❄️ Hi, I'm the [Snowflake Cortex Analyst](https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-analyst)!
I can answer questions about the `{model_file}` data model or generate SQL
queries for you."
))
observeEvent(input$chat_user_input, {
chat_append("chat", chat$stream_async(input$chat_user_input))
})
}
shinyApp(ui, server)
Unit tests that cover basic functionality are included, but there is no
integration test per se that uses a real Connect server.
Signed-off-by: Aaron Jacobs <aaron.jacobs@posit.co>
|
Superseded by #183. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Note: this is stacked on top of #145. Ignore the first commit.
This commit ports
odbc's recently-added support for "viewer-based credentials" on Posit Connect toelmer. Specifically, passing a Shiny session object tochat_cortex()now allows Cortex chatbots to use the viewer's Snowflake credentials on Connect rather than requiring a service account or the publisher to embed their personal credentials.For example:
Unit tests that cover basic functionality are included, but there is no integration test per se that uses a real Connect server.