mirror of
https://github.com/discourse/discourse.git
synced 2026-08-05 11:15:10 +08:00
86680f6b0dFollow up to https://github.com/discourse/discourse/pull/41586 1. **Bulk of the PR diff** -- there's a mix of usages for "description" and "caption", we'll stick to V1 terminology and keep "AI caption" instead of bringing in "description". 2. Move image caption out of AI helper, into its own dedicated "AI Feature" for consolidated settings 3. Update default values for the existing image caption agent - vision_enabled was default off, now on - update description 4. Validations and problem checks - via Site Setting validation - agent / llm - when enabling the feature - or when selecting invalid agent - inform admin when they've enabled the setting but the agent is misconfigured via ProblemCheck
33 lines
738 B
Ruby
Vendored
33 lines
738 B
Ruby
Vendored
# frozen_string_literal: true
|
|
|
|
module DiscourseAi
|
|
module Agents
|
|
class ImageCaptioner < Agent
|
|
def self.default_enabled
|
|
false
|
|
end
|
|
|
|
def self.vision_enabled
|
|
true
|
|
end
|
|
|
|
def system_prompt
|
|
<<~PROMPT.strip
|
|
You are a bot specializing in image captioning.
|
|
|
|
Format your response as a JSON object with a single key named "output", which has the caption as the value.
|
|
Your output should be in the following format:
|
|
|
|
{"output": "xx"}
|
|
|
|
Where "xx" is replaced by the caption.
|
|
reply with valid JSON only
|
|
PROMPT
|
|
end
|
|
|
|
def response_format
|
|
[{ "key" => "output", "type" => "string" }]
|
|
end
|
|
end
|
|
end
|
|
end
|
