Learn how Action Nodes in Giselle enable your workflows to interact with external services like GitHub, performing tasks such as creating issues, comments, or retrieving discussions in your repositories.
giselles-ai
, liam-hq
).docs
, giselle
).giselle
).giselles-ai/giselle
), you’ll need to specify what action the Giselle’s app should execute.issueNumber
and a body
for the comment, as indicated by “Required” fields.issueNumber
and body
).giselles-ai/giselle
repository:
giselles-ai/giselle
issueNumber
: Sourced from a workflow trigger or a previous node (e.g., an issue number identified earlier in the flow).body
: Sourced from another node (e.g., a text summary generated by an AI model in a preceding step).body
to the issue specified by issueNumber
within the giselles-ai/giselle
repository.giselles-ai/giselle
repository, as shown in the provided image:
giselles-ai/giselle
pullNumber
: The number of the pull request containing the review comment. This would typically be sourced from a trigger event (e.g., a GitHub Trigger for “Pull Request Review Comment Created”) or a previous node that identified the pull request.commentId
: The ID of the specific review comment to which you are replying. This also would likely come from the trigger data or a preceding node that fetched comment details.body
: The content of your reply. This could be static text, or more powerfully, dynamically generated text from an AI node in your workflow (e.g., an automated acknowledgment, a question, or a summary).body
parameter as a reply to the specified commentId
on the given pullNumber
within the giselles-ai/giselle
repository.giselles-ai/giselle
(or any other configured repository)discussionNumber
: The number of the GitHub Discussion to retrieve. This can be a static value or sourced dynamically from a previous node’s output or a trigger input.discussionNumber
from the selected repository.id
, title
, bodyHTML
(the content of the discussion in HTML format), createdAt
, updatedAt
, author
information, and an array of comments
associated with the discussion (each comment also having details like bodyHTML
, author
, etc.). This structured data can then be used by subsequent nodes in your workflow, for example, to summarize the discussion, extract key points, perform analysis on the comments, or trigger other actions based on its content.