Compare commits

...
Author SHA1 Message Date
Keegan George 9c80396942 DEV: Re-introduce validator 2025-06-03 06:30:50 -07:00
Sam 4dffd0b2c5 DEV: improve tool infra, improve forum researcher prompts, improve logging (#1391)
- add sleep function for tool polling with rate limits
- Support base64 encoding for HTTP requests and uploads
-  Enhance forum researcher with cost warnings and comprehensive planning
- Add cancellation support for research operations
- Include feature_name parameter for bot analytics
- richer research support (OR queries)
2025-06-03 15:17:55 +10:00
Keegan George 4c0660d6fd DEV: Remove validator for spam (#1393) 2025-06-02 13:57:34 -07:00
Rafael dos Santos Silva 27de71fc4f FIX: Proper default LLM detection for inferred concepts (#1392) 2025-06-02 17:56:47 -03:00
478f31de47 FEATURE: add inferred concepts system (#1330)
* FEATURE: add inferred concepts system

This commit adds a new inferred concepts system that:
- Creates a model for storing concept labels that can be applied to topics
- Provides AI personas for finding new concepts and matching existing ones
- Adds jobs for generating concepts from popular topics
- Includes a scheduled job that automatically processes engaging topics

* FEATURE: Extend inferred concepts to include posts

* Adds support for concepts to be inferred from and applied to posts
* Replaces daily task with one that handles both topics and posts
* Adds database migration for posts_inferred_concepts join table
* Updates PersonaContext to include inferred concepts



Co-authored-by: Roman Rizzi <[email protected]>
Co-authored-by: Keegan George <[email protected]>
2025-06-02 14:29:20 -03:00
David Taylor 4ce8973e56 PERF: Optimize .ai-debug-modal__tokens selector (#1390)
This is showing as the most expensive CSS selector in Discourse at the moment. Adding specific classes and dropping the general `span` selector will make this much cheaper.
2025-05-30 21:47:30 +01:00
Keegan George 34c98de864 FIX: Exporting overall sentiment fails (#1388)
## 🔍 Overview

When exporting an Overall Sentiment report in the admin panel, the export fails with:

```ruby
Job exception: no implicit conversion of Symbol into Integer
```

This was happening because we are passing a single _Hash_ to `report.data` however, exports expect `report.data` to be an _Array of Hashes_. This update fixes this issue by wrapping the data in an array.
2025-05-30 11:58:28 -07:00
Keegan George 38f7e9c2c4 UX: AI composer helper refinements (#1387)
This update includes a variety of small refinements to the AI composer helper:

- prevent height jump when going from loading text placeholder → proofreading text streaming
- update padding on AI helper options list to be more suitable with typical Discourse menu design
- for composer helper results that are not `showResultAsDiff` (i.e. translation):
   - update before/after diff design to be more subtle
   - results should be in normal font (as the text is cooked and not raw markdown)
- fix: smooth streaming animation stuck showing dot icon even after smooth streaming is done
2025-05-30 10:35:53 -07:00
Sam e6876aabd5 FIX: enum handling needs to be done on save as well (#1386) 2025-05-30 17:58:16 +10:00
Sam b5d393b4bc FIX: custom tools incorrectly setting all fields to blank enum (#1385)
Previous to this change, enum was set to [] which broke all non
enum tools
2025-05-30 17:12:24 +10:00
Sam 77ae426d95 FEATURE: support upload.getUrl in custom tools (#1384)
* FEATURE: support upload.getUrl in custom tools

Some tools need to share images with an API. A common pattern
is for APIs to expect a URL.

This allows converting upload://123123 to a proper CDN friendly
URL from within a custom tool

* no support for secure uploads, so be explicit about it.
2025-05-30 15:47:07 +10:00
Alan Guo Xiang Tan 9f43df0302 FIX: Full page search broken (#1383)
Follow-up to 59d6e2b467
2025-05-30 10:07:48 +08:00
David Taylor 59d6e2b467 DEV: Replace narrowDesktopView logic with viewport API (#1372) 2025-05-30 06:52:39 +10:00
SamandKeegan George c06d7b07d5 FEATURE: simplify streaming implementation - rush last update (#1380)
* FEATURE: simplify streaming implementation - rush last update

Previous to this change we would simply "flash" the final update
on the screen, this amends it so we quickly update the UI in about
1 second in the end with all the final update.

This makes the UI feel more interactive to end users.

* DEV: Updates...

- Remove unnecessary comments
- Add doc style comments for all methods
- Organize methods (private at bottom)
- Update some variable names

---------

Co-authored-by: Keegan George <[email protected]>
2025-05-30 06:50:12 +10:00
Natalie Tay 373e2305d6 FEATURE: Automatic translation and localization of posts, topics, categories (#1376)
Related: https://github.com/discourse/discourse-translator/pull/310

This commit includes all the jobs and event hooks to localize posts, topics, and categories.

A few notes:
- `feature_name: "translation"` because the site setting is `ai-translation` and module is `Translation`
- we will switch to proper ai-feature in the near future, and can consider using the persona_user as `localization.localizer_user_id`
- keeping things flat within the module for now as we will be moving to ai-feature soon and have to rearrange
- Settings renamed/introduced are:
  - ai_translation_backfill_rate (0)
  - ai_translation_backfill_limit_to_public_content (true)
  - ai_translation_backfill_max_age_days (5)
  - ai_translation_verbose_logs (false)
2025-05-29 17:28:06 +08:00
Guhyoun Nam ad5c48d9ae DEV: Add appEvents trigger for AI New Question button (#1379)
* DEV: Add appEvents trigger for AI New Question button

* appEvent name update
2025-05-28 13:26:37 -05:00
Keegan George d99c335dab DEV: Ensure enabling/disabling spam is set and logged (#1378)
Since we enable/disable `ai_spam_detection_enabled` setting in a custom Spam tab UI in AI, we want to ensure we retain the setting and logging features. To preserve that, we want to update the controller to use `SiteSetting.set_and_log` instead of setting the value directly.
2025-05-28 10:12:21 -07:00
Roman Rizzi 01e29ca5d8 FIX: Bump persona's examples length (#1377) 2025-05-28 14:01:44 -03:00
Keegan George 297c64c3b8 DEV: Unhide spam detection setting (#1374)
## 🔍 Overview
We want to unhide `ai_spam_detection_enabled` setting so that we can retain staff action log features. However, we also want to ensure users cannot enable spam detection without having `AiModerationSetting.spam` present in the database.

In this update we unhide the setting, but also add a validator to ensure the necessary configuration is in place before allowing the setting to be enabled.
2025-05-28 07:50:56 -07:00
David Taylor ca78b1a1c5 DEV: Patch Net::BufferedIO to help debug spec flakes (#1375)
Internal `/t/154170`
2025-05-28 10:24:07 +01:00
Discourse Translator Bot 39653aed22 Update translations (#1371) 2025-05-27 22:11:32 +02:00
Sam 6781ecd02e DEV: cleanup diff streaming (#1370)
This simplifies some of the internal logic and ensures it is a
bit more robust
2025-05-27 18:12:02 +10:00
Keegan George e264572597 FIX: Closing AI menu stops post audio from playing (#1369)
## 🔍 Overview
When you have a post with audio being played and you open and close the AI post helper menu, it re-renders the entire post DOM, causing the audio to be interrupted and stop playing.

The reason for this is because we highlight the selected text when opening the AI post helper menu and we replace the entire post back with the original post HTML when closing the menu. This fix ensures that we do not re-render the entire post DOM and instead only remove the highlighted section that was added.

## 🔗 Context
https://meta.discourse.org/t/ai-helper-interrupting-uploaded-mp3-audio-stream/366817?u=keegan
2025-05-26 10:24:06 -07:00
Sam 70b0db2871 FIX: improve researcher tool - fix topic filters (#1368)
* Small fix, reasoning is now available on Claude 4 models

* fix invalid filters should raise, topic filter not working

* fix spec so we are consistent
2025-05-26 16:00:44 +10:00
Sam 2d6ec5e1e6 FIX: apply diffs more consistently (#1367)
* FIX: apply diffs more consistently

1. Do escaping direct in diff streamer, that way HTML tags and other unsafe chars can be displayed and fixed
2. Add safeguard to ensure streaming always stops when it was terminated elsewhere

* lint

* bug unsubscribe should unsubscribe
2025-05-24 15:19:48 +10:00
Mark VanLandingham cead887480 FIX: Don't error when navigating from AI Bot topic to regular (#1366)
We were getting an error in this logic causing Ember to fail to render the non-bot-topic that we navigate to.

I believe this is because the getter of participants is re-calculating (due to this.header.topicInfo being updated) before the args to this connector changes. Adding some safe navigation here fixes the issue.
2025-05-23 13:30:08 -05:00
Jordan Vidrine adc92e1e75 UX: Allow rounded corners (#1365) 2025-05-23 10:42:55 -05:00
Roman Rizzi 0ce17a122f FIX: Correctly pass tool_choice when using Claude models. (#1364)
The `ClaudePrompt` object couldn't access the original prompt's tool_choice attribute, affecting both Anthropic and Bedrock.
2025-05-23 10:36:52 -03:00
Sam cf220c530c FIX: Improve MessageBus efficiency and correctly stop streaming (#1362)
* FIX: Improve MessageBus efficiency and correctly stop streaming

This commit enhances the message bus implementation for AI helper streaming by:

- Adding client_id targeting for message bus publications to ensure only the requesting client receives streaming updates
- Limiting MessageBus backlog size (2) and age (60 seconds) to prevent Redis bloat
- Replacing clearTimeout with Ember's cancel method for proper runloop management, we were leaking a stop
- Adding tests for client-specific message delivery

These changes improve memory usage and make streaming more reliable by ensuring messages are properly directed to the requesting client.

* composer suggestion needed a fix as well.

* backlog size of 2 is risky here cause same channel name is reused between clients
2025-05-23 16:23:06 +10:00
Keegan George 61ef1932fa DEV: Follow-up small fixes to AI composer helper (#1361)
**This update includes a variety of small fixes to the AI composer helper:**
- ensures there is no flash of no text when diff modal is triggered
- escapes selected text prior to diff streaming
- uses monospace font in diff modal since text rendered is raw markdown
2025-05-22 12:37:42 -07:00
Keegan George 11e90f5f3f DEV: Hide proofreading shortcut in toolbar title (#1360) 2025-05-22 09:36:55 -07:00
Joffrey JAFFEUX b817810d46 FIX: ensures stream update object is scoped to its initial topic (#1359)
* FIX: ensures stream update object is scoped to its initial topic

Before this commit you could end up in this situation where a `post-updater` is constructed for a specific topic, but the user changes topic mid steam and it ends up updating the same post number but in a  different topic as we were only checking for `post_number` and not the combination of `topic_id` + `post_number`.

* we should have the topic
2025-05-22 20:08:43 +10:00
Sam 55dab9c68a FIX: stabilize diff algorithm for streaming (#1358)
Previous to this fix we would diff the entire body of text, this
could lead to situations where a diff presented to a user was
wildly off matching areas in the text that should not have been
tested.

New algorithm only checks a portion of the string, this ensures
that during streaming there is no chance for wild mistakes
2025-05-22 16:31:21 +10:00
Keegan GeorgeandSam Saffron c29183fc2d FEATURE: Add diff streaming animation (#1355)
Previously we attempted to add a diff streaming animation to the AI composer helper: https://github.com/discourse/discourse-ai/pull/1332, but it resulted in issues despite attempted fixes (https://github.com/discourse/discourse-ai/pull/1338) so we temporarily suppressed the diff animation (https://github.com/discourse/discourse-ai/pull/1341).

This update makes a second attempt at implementing the diff streaming animation. Instead of creating a custom diff algorithm, we make use of a third-party library [`jsDiff`](https://github.com/kpdecker/jsdiff) (which we added to core here: https://github.com/discourse/discourse/pull/32833). While streaming, the diff animation often struggles with markdown links and images, so we make use of `markdown-it` parser to detect those cases and prevent breaking the animation.

---------

Co-authored-by: Sam Saffron <[email protected]>
2025-05-22 08:10:50 +10:00
Roman Rizzi d72ad84f8f FIX: Retry parsing escaped inner JSON to handle control chars. (#1357)
The structured output JSON comes embedded inside the API response, which is also a JSON. Since we have to parse the response to process it, any control characters inside the structured output are unescaped into regular characters, leading to invalid JSON and breaking during parsing. This change adds a retry mechanism that escapes
the string again if parsing fails, preventing the parser from breaking on malformed input and working around this issue.

For example:

```
  original = '{ "a": "{\\"key\\":\\"value with \\n newline\\"}" }'
  JSON.parse(original) => { "a" => "{\"key\":\"value with \n newline\"}" }
  # At this point, the inner JSON string contains an actual newline.
```
2025-05-21 11:25:59 -03:00
Roman Rizzi e207eba1a4 FIX: Don't dig on nil when checking for the gemini schema (#1356) 2025-05-21 08:30:47 -03:00
Kris 53905f65ac FEATURE: add participants and invite button to AI conversations (#1354) 2025-05-21 10:27:06 +10:00
Guhyoun Nam ef260be7cf DEV: Move the stop-forcing-conversations-sidebar appEvents trigger (#1353) 2025-05-20 15:06:54 -05:00
Discourse Translator Bot 0c69bb85ce Update translations (#1352) 2025-05-20 17:33:52 +02:00
Sam af18603b21 DEV: cancel manager should bypass webmock (#1350)
Webmock can be a bit flaky under certain use cases.
2025-05-20 13:01:55 +10:00
Sam 7db2589cc4 DEV: prompt engineering to improve citations (#1351) 2025-05-20 13:01:35 +10:00
Roman Rizzi 2fb691cba8 FEATURE: Triage can hide posts after adding them to the review queue (#1348) 2025-05-20 08:19:00 +10:00
Sam de0625571b FIX: add safe navigation to serializer include conditions (#1349)
In some rare cases a post can exist without a topic
2025-05-19 18:55:55 -03:00
Joffrey JAFFEUX 296aa24df1 DEV: rewrites artifact spec with capybara waiters (#1347)
Generally speaking we never want to do:

```
expect(element.text).to eq("foo")
```

As these are rspec matchers and do not add further Capybara-style waiting specifically for the text content to become present.
2025-05-20 07:27:15 +10:00
Sam 3ac2359ff1 FEATURE: allow passing in data attributes to an artifact (#1346)
Also allow artifact access to current username

Usage inside artifact is:

1. await window.discourseArtifactReady;
2. access data via window.discourseArtifactData;
2025-05-19 15:44:37 +10:00
Guhyoun Nam 925949de47 DEV: Add multiple appEvents trigger on AI Chatbot (#1324)
* DEV: Add appEvents trigger on AI Chatbot header icon

* update event name

* stop-forcing-custom-sidebar appevents trigger

* add force stop custom sidebar appEvents

* trigger name update

* trigger name update
2025-05-16 12:25:16 -05:00
David Taylor b2fbab9fad DEV: Fix linting violations (#1345) 2025-05-16 13:06:15 +01:00
David Taylor 381fa14158 DEV: Rename spec (#1344)
This has nothing to do with `assets:precompile`. Likely a copy/paste from another spec
2025-05-16 09:40:08 +01:00
David Taylor 77d3a38e49 FIX: AI share page assets via CDN on login-required sites (#1343)
AI share page assets are loaded via the app CDN, which means the requests have no authentication and will never appear to the app as "logged in". Therefore we should skip the `redirect_to_login_if_required` before_action.
2025-05-16 09:30:38 +01:00
Sam 7316058dfc FEATURE: hashtag and mention autocomplete for first bot message (#1342)
Also removes controller which is a deprecated pattern

* some comment improvements

* remove uneeded code
2025-05-16 16:19:20 +10:00
Keegan George 9ee82fd8be DEV: Temporarily suppress diff animation as we fix issues (#1341)
The diff animation introduced in https://github.com/discourse/discourse-ai/pull/1332 and with attempts to improve it in https://github.com/discourse/discourse-ai/pull/1338 still has various issues. As we work on a fix, we want to revert the animation to simply stream the diff without animation so users are not left with a janky unusable experience.
2025-05-15 14:55:30 -07:00
Keegan George dfea784fc4 DEV: Improve diff streaming accuracy with safety checker (#1338)
This update adds a safety checker which scans the streamed updates. It ensures that incomplete segments of text are not sent yet over message bus as this will cause breakage with the diff streamer. It also updates the diff streamer to handle a thinking state for when we are waiting for message bus updates.
2025-05-15 11:38:46 -07:00
Roman Rizzi ff2e18f9ca FIX: Structured output discrepancies. (#1340)
This change fixes two bugs and adds a safeguard.

The first issue is that the schema Gemini expected differed from the one sent, resulting in 400 errors when performing completions.

The second issue was that creating a new persona won't define a method
for `response_format`. This has to be explicitly defined when we wrap it inside the Persona class. Also, There was a mismatch between the default value and what we stored in the DB. Some parts of the code expected symbols as keys and others as strings.

Finally, we add a safeguard when, even if asked to, the model refuses to reply with a valid JSON. In this case, we are making a best-effort to recover and stream the raw response.
2025-05-15 11:32:10 -03:00
Sam 1b3fdad5c7 FEATURE: allow researcher to also research specific topics (#1339)
* FEATURE: allow researcher to also research specific topics

Also improve UI around research with more accurate info

* this ensures that under no conditions PMs will be included
2025-05-15 17:48:21 +10:00
SamandCopilot 2c6459429f DEV: use a proper object for tool definition (#1337)
* DEV: use a proper object for tool definition

This moves away from using a loose hash to define tools, which
is error prone.

Instead given a proper object we will also be able to coerce the
return values to match tool definition correctly

* fix xml tools

* fix anthropic tools

* fix specs... a few more to go

* specs are passing

* FIX: coerce values for XML tool calls

* Update spec/lib/completions/tool_definition_spec.rb

Co-authored-by: Copilot <[email protected]>

---------

Co-authored-by: Copilot <[email protected]>
2025-05-15 17:32:39 +10:00
Sam c34fcc8a95 FEATURE: forum researcher persona for deep research (#1313)
This commit introduces a new Forum Researcher persona specialized in deep forum content analysis along with comprehensive improvements to our AI infrastructure.

Key additions:

    New Forum Researcher persona with advanced filtering and analysis capabilities
    Robust filtering system supporting tags, categories, dates, users, and keywords
    LLM formatter to efficiently process and chunk research results

Infrastructure improvements:

    Implemented CancelManager class to centrally manage AI completion cancellations
    Replaced callback-based cancellation with a more robust pattern
    Added systematic cancellation monitoring with callbacks

Other improvements:

    Added configurable default_enabled flag to control which personas are enabled by default
    Updated translation strings for the new researcher functionality
    Added comprehensive specs for the new components

    Renames Researcher -> Web Researcher

This change makes our AI platform more stable while adding powerful research capabilities that can analyze forum trends and surface relevant content.
2025-05-14 12:36:16 +10:00
257 changed files with 10601 additions and 997 deletions
@@ -50,7 +50,7 @@ module DiscourseAi
)
end
SiteSetting.ai_spam_detection_enabled = is_enabled
SiteSetting.set_and_log("ai_spam_detection_enabled", is_enabled, current_user)
end
render json: AiSpamSerializer.new(spam_config, root: false)
@@ -42,6 +42,21 @@ module DiscourseAi
<style>
#{artifact.css}
</style>
<script>
window._discourse_user_data = {
#{current_user ? "username: #{current_user.username.to_json}" : "username: null"}
};
window.discourseArtifactReady = new Promise(resolve => {
window._resolveArtifactData = resolve;
});
window.addEventListener('message', function(event) {
if (event.data && event.data.type === 'discourse-artifact-data') {
window.discourseArtifactData = event.data.dataset || {};
Object.assign(window.discourseArtifactData, window._discourse_user_data);
window._resolveArtifactData(window.discourseArtifactData);
}
});
</script>
</head>
<body>
#{artifact.html}
@@ -74,6 +89,19 @@ module DiscourseAi
</head>
<body>
<iframe sandbox="allow-scripts allow-forms" height="100%" width="100%" srcdoc="#{ERB::Util.html_escape(untrusted_html)}" frameborder="0"></iframe>
<script>
document.querySelector('iframe').addEventListener('load', function() {
try {
const iframeWindow = this.contentWindow;
const message = { type: 'discourse-artifact-data', dataset: {} };
if (window.frameElement && window.frameElement.dataset) {
Object.assign(message.dataset, window.frameElement.dataset);
}
iframeWindow.postMessage(message, '*');
} catch (e) { console.error('Error passing data to artifact:', e); }
});
</script>
</body>
</html>
HTML
@@ -8,7 +8,7 @@ module DiscourseAi
before_action :require_site_settings!
skip_before_action :preload_json, :check_xhr, only: %i[show asset]
skip_before_action :verify_authenticity_token, only: ["asset"]
skip_before_action :redirect_to_login_if_required, :verify_authenticity_token, only: %i[asset]
def create
ensure_allowed_create!
@@ -124,6 +124,10 @@ module DiscourseAi
raise Discourse::InvalidParameters.new(:custom_prompt) if params[:custom_prompt].blank?
end
# to stream we must have an appropriate client_id
# otherwise we may end up streaming the data to the wrong client
raise Discourse::InvalidParameters.new(:client_id) if params[:client_id].blank?
if location == "composer"
Jobs.enqueue(
:stream_composer_helper,
@@ -132,6 +136,7 @@ module DiscourseAi
prompt: prompt.name,
custom_prompt: params[:custom_prompt],
force_default_locale: params[:force_default_locale] || false,
client_id: params[:client_id],
)
else
post_id = get_post_param!
@@ -146,6 +151,7 @@ module DiscourseAi
text: text,
prompt: prompt.name,
custom_prompt: params[:custom_prompt],
client_id: params[:client_id],
)
end
+1 -1
View File
@@ -15,7 +15,7 @@ module ::Jobs
bot = DiscourseAi::Personas::Bot.as(bot_user, persona: persona.new)
DiscourseAi::AiBot::Playground.new(bot).reply_to(post)
DiscourseAi::AiBot::Playground.new(bot).reply_to(post, feature_name: "bot")
rescue DiscourseAi::Personas::Bot::BOT_NOT_FOUND
Rails.logger.warn(
"Bot not found for post #{post.id} - perhaps persona was deleted or bot was disabled",
+51
View File
@@ -0,0 +1,51 @@
# frozen_string_literal: true
module Jobs
class DetectTranslatePost < ::Jobs::Base
sidekiq_options retry: false
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
return if args[:post_id].blank?
post = Post.find_by(id: args[:post_id])
return if post.blank? || post.raw.blank? || post.deleted_at.present? || post.user_id <= 0
if SiteSetting.ai_translation_backfill_limit_to_public_content
topic = post.topic
if topic.blank? || topic.category&.read_restricted? ||
topic.archetype == Archetype.private_message
return
end
end
begin
detected_locale = DiscourseAi::Translation::PostLocaleDetector.detect_locale(post)
rescue FinalDestination::SSRFDetector::LookupFailedError
# this job is non-critical
# the backfill job will handle failures
return
end
locales = SiteSetting.experimental_content_localization_supported_locales.split("|")
return if locales.blank?
locales.each do |locale|
next if locale == detected_locale
begin
DiscourseAi::Translation::PostLocalizer.localize(post, locale)
rescue FinalDestination::SSRFDetector::LookupFailedError
# do nothing, there are too many sporadic lookup failures
rescue => e
DiscourseAi::Translation::VerboseLogger.log(
"Failed to translate post #{post.id} to #{locale}: #{e.message}",
)
end
end
MessageBus.publish("/topic/#{post.topic_id}", type: :localized, id: post.id)
end
end
end
@@ -0,0 +1,49 @@
# frozen_string_literal: true
module Jobs
class DetectTranslateTopic < ::Jobs::Base
sidekiq_options retry: false
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
return if args[:topic_id].blank?
topic = Topic.find_by(id: args[:topic_id])
if topic.blank? || topic.title.blank? || topic.deleted_at.present? || topic.user_id <= 0
return
end
if SiteSetting.ai_translation_backfill_limit_to_public_content
return if topic.category&.read_restricted?
end
begin
detected_locale = DiscourseAi::Translation::TopicLocaleDetector.detect_locale(topic)
rescue FinalDestination::SSRFDetector::LookupFailedError
# this job is non-critical
# the backfill job will handle failures
return
end
locales = SiteSetting.experimental_content_localization_supported_locales.split("|")
return if locales.blank?
locales.each do |locale|
next if locale == detected_locale
begin
DiscourseAi::Translation::TopicLocalizer.localize(topic, locale)
rescue FinalDestination::SSRFDetector::LookupFailedError
# do nothing, there are too many sporadic lookup failures
rescue => e
DiscourseAi::Translation::VerboseLogger.log(
"Failed to translate topic #{topic.id} to #{locale}: #{e.message}",
)
end
end
MessageBus.publish("/topic/#{topic.id}", type: :localized, id: 1)
end
end
end
@@ -0,0 +1,70 @@
# frozen_string_literal: true
module Jobs
class GenerateInferredConcepts < ::Jobs::Base
sidekiq_options queue: "low"
# Process items to generate new concepts
#
# @param args [Hash] Contains job arguments
# @option args [String] :item_type Required - Type of items to process ('topics' or 'posts')
# @option args [Array<Integer>] :item_ids Required - List of item IDs to process
# @option args [Integer] :batch_size (100) Number of items to process in each batch
# @option args [Boolean] :match_only (false) Only match against existing concepts without generating new ones
def execute(args = {})
return if args[:item_ids].blank? || args[:item_type].blank?
if %w[topics posts].exclude?(args[:item_type])
Rails.logger.error("Invalid item_type for GenerateInferredConcepts: #{args[:item_type]}")
return
end
# Process items in smaller batches to avoid memory issues
batch_size = args[:batch_size] || 100
# Get the list of item IDs
item_ids = args[:item_ids]
match_only = args[:match_only] || false
# Process items in batches
item_ids.each_slice(batch_size) do |batch_item_ids|
process_batch(batch_item_ids, args[:item_type], match_only)
end
end
private
def process_batch(item_ids, item_type, match_only)
klass = item_type.singularize.classify.constantize
items = klass.where(id: item_ids)
manager = DiscourseAi::InferredConcepts::Manager.new
items.each do |item|
begin
process_item(item, item_type, match_only, manager)
rescue => e
Rails.logger.error(
"Error generating concepts from #{item_type.singularize} #{item.id}: #{e.message}\n#{e.backtrace.join("\n")}",
)
end
end
end
def process_item(item, item_type, match_only, manager)
# Use the Manager method that handles both identifying and creating concepts
if match_only
if item_type == "topics"
manager.match_topic_to_concepts(item)
else # posts
manager.match_post_to_concepts(item)
end
else
if item_type == "topics"
manager.generate_concepts_from_topic(item)
else # posts
manager.generate_concepts_from_post(item)
end
end
end
end
end
+51
View File
@@ -0,0 +1,51 @@
# frozen_string_literal: true
module Jobs
class LocalizeCategories < ::Jobs::Base
cluster_concurrency 1
sidekiq_options retry: false
BATCH_SIZE = 50
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
locales = SiteSetting.experimental_content_localization_supported_locales.split("|")
return if locales.blank?
cat_id = args[:from_category_id] || Category.order(:id).first&.id
last_id = nil
categories = Category.where("id >= ?", cat_id).order(:id).limit(BATCH_SIZE)
return if categories.empty?
categories.each do |category|
if SiteSetting.ai_translation_backfill_limit_to_public_content && category.read_restricted?
last_id = category.id
next
end
CategoryLocalization.transaction do
locales.each do |locale|
next if CategoryLocalization.exists?(category_id: category.id, locale: locale)
begin
DiscourseAi::Translation::CategoryLocalizer.localize(category, locale)
rescue FinalDestination::SSRFDetector::LookupFailedError
# do nothing, there are too many sporadic lookup failures
rescue => e
DiscourseAi::Translation::VerboseLogger.log(
"Failed to translate category #{category.id} to #{locale}: #{e.message}",
)
end
end
end
last_id = category.id
end
if categories.size == BATCH_SIZE
Jobs.enqueue_in(10.seconds, :localize_categories, from_category_id: last_id + 1)
end
end
end
end
+68
View File
@@ -0,0 +1,68 @@
# frozen_string_literal: true
module Jobs
class LocalizePosts < ::Jobs::Base
cluster_concurrency 1
sidekiq_options retry: false
BATCH_SIZE = 50
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
locales = SiteSetting.experimental_content_localization_supported_locales.split("|")
return if locales.blank?
limit = args[:limit] || BATCH_SIZE
locales.each do |locale|
posts =
Post
.joins(
"LEFT JOIN post_localizations pl ON pl.post_id = posts.id AND pl.locale = #{ActiveRecord::Base.connection.quote(locale)}",
)
.where(deleted_at: nil)
.where("posts.user_id > 0")
.where.not(raw: [nil, ""])
.where.not(locale: nil)
.where.not(locale: locale)
.where("pl.id IS NULL")
if SiteSetting.ai_translation_backfill_limit_to_public_content
posts =
posts
.joins(:topic)
.where(topics: { category_id: Category.where(read_restricted: false).select(:id) })
.where.not(topics: { archetype: Archetype.private_message })
end
if SiteSetting.ai_translation_backfill_max_age_days > 0
posts =
posts.where(
"posts.created_at > ?",
SiteSetting.ai_translation_backfill_max_age_days.days.ago,
)
end
posts = posts.order(updated_at: :desc).limit(limit)
next if posts.empty?
posts.each do |post|
begin
DiscourseAi::Translation::PostLocalizer.localize(post, locale)
rescue FinalDestination::SSRFDetector::LookupFailedError
# do nothing, there are too many sporadic lookup failures
rescue => e
DiscourseAi::Translation::VerboseLogger.log(
"Failed to translate post #{post.id} to #{locale}: #{e.message}",
)
end
end
DiscourseAi::Translation::VerboseLogger.log("Translated #{posts.size} posts to #{locale}")
end
end
end
end
+63
View File
@@ -0,0 +1,63 @@
# frozen_string_literal: true
module Jobs
class LocalizeTopics < ::Jobs::Base
cluster_concurrency 1
sidekiq_options retry: false
BATCH_SIZE = 50
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
locales = SiteSetting.experimental_content_localization_supported_locales.split("|")
return if locales.blank?
limit = args[:limit] || BATCH_SIZE
locales.each do |locale|
topics =
Topic
.joins(
"LEFT JOIN topic_localizations tl ON tl.topic_id = topics.id AND tl.locale = #{ActiveRecord::Base.connection.quote(locale)}",
)
.where(deleted_at: nil)
.where("topics.user_id > 0")
.where.not(locale: nil)
.where.not(locale: locale)
.where("tl.id IS NULL")
if SiteSetting.ai_translation_backfill_limit_to_public_content
topics = topics.where(category_id: Category.where(read_restricted: false).select(:id))
end
if SiteSetting.ai_translation_backfill_max_age_days > 0
topics =
topics.where(
"topics.created_at > ?",
SiteSetting.ai_translation_backfill_max_age_days.days.ago,
)
end
topics = topics.order(updated_at: :desc).limit(limit)
next if topics.empty?
topics.each do |topic|
begin
DiscourseAi::Translation::TopicLocalizer.localize(topic, locale)
rescue FinalDestination::SSRFDetector::LookupFailedError
# do nothing, there are too many sporadic lookup failures
rescue => e
DiscourseAi::Translation::VerboseLogger.log(
"Failed to translate topic #{topic.id} to #{locale}: #{e.message}",
)
end
end
DiscourseAi::Translation::VerboseLogger.log("Translated #{topics.size} topics to #{locale}")
end
end
end
end
@@ -8,6 +8,7 @@ module Jobs
return unless args[:prompt]
return unless user = User.find_by(id: args[:user_id])
return unless args[:text]
return unless args[:client_id]
prompt = CompletionPrompt.enabled_by_name(args[:prompt])
@@ -21,6 +22,7 @@ module Jobs
user,
"/discourse-ai/ai-helper/stream_composer_suggestion",
force_default_locale: args[:force_default_locale],
client_id: args[:client_id],
)
end
end
@@ -0,0 +1,16 @@
# frozen_string_literal: true
module Jobs
class CategoryLocalizationBackfill < ::Jobs::Scheduled
every 12.hours
cluster_concurrency 1
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
return if SiteSetting.experimental_content_localization_supported_locales.blank?
Jobs.enqueue(:localize_categories)
end
end
end
@@ -0,0 +1,87 @@
# frozen_string_literal: true
module Jobs
class GenerateConceptsFromPopularItems < ::Jobs::Scheduled
every 1.day
# This job runs daily and generates new concepts from popular topics and posts
# It selects items based on engagement metrics and generates concepts from their content
def execute(_args)
return unless SiteSetting.inferred_concepts_enabled
process_popular_topics
process_popular_posts
end
private
def process_popular_topics
# Find candidate topics that are popular and don't have concepts yet
manager = DiscourseAi::InferredConcepts::Manager.new
candidates =
manager.find_candidate_topics(
limit: SiteSetting.inferred_concepts_daily_topics_limit || 20,
min_posts: SiteSetting.inferred_concepts_min_posts || 5,
min_likes: SiteSetting.inferred_concepts_min_likes || 10,
min_views: SiteSetting.inferred_concepts_min_views || 100,
created_after: SiteSetting.inferred_concepts_lookback_days.days.ago,
)
return if candidates.blank?
# Process candidate topics - first generate concepts, then match
Jobs.enqueue(
:generate_inferred_concepts,
item_type: "topics",
item_ids: candidates.map(&:id),
batch_size: 10,
)
if SiteSetting.inferred_concepts_background_match
# Schedule a follow-up job to match existing concepts
Jobs.enqueue_in(
1.hour,
:generate_inferred_concepts,
item_type: "topics",
item_ids: candidates.map(&:id),
batch_size: 10,
match_only: true,
)
end
end
def process_popular_posts
# Find candidate posts that are popular and don't have concepts yet
manager = DiscourseAi::InferredConcepts::Manager.new
candidates =
manager.find_candidate_posts(
limit: SiteSetting.inferred_concepts_daily_posts_limit || 30,
min_likes: SiteSetting.inferred_concepts_post_min_likes || 5,
exclude_first_posts: true,
created_after: SiteSetting.inferred_concepts_lookback_days.days.ago,
)
return if candidates.blank?
# Process candidate posts - first generate concepts, then match
Jobs.enqueue(
:generate_inferred_concepts,
item_type: "posts",
item_ids: candidates.map(&:id),
batch_size: 10,
)
if SiteSetting.inferred_concepts_background_match
# Schedule a follow-up job to match against existing concepts
Jobs.enqueue_in(
1.hour,
:generate_inferred_concepts,
item_type: "posts",
item_ids: candidates.map(&:id),
batch_size: 10,
match_only: true,
)
end
end
end
end
@@ -0,0 +1,18 @@
# frozen_string_literal: true
module Jobs
class PostLocalizationBackfill < ::Jobs::Scheduled
every 5.minutes
cluster_concurrency 1
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
return if SiteSetting.experimental_content_localization_supported_locales.blank?
return if SiteSetting.ai_translation_backfill_rate == 0
Jobs.enqueue(:localize_posts, limit: SiteSetting.ai_translation_backfill_rate)
end
end
end
@@ -0,0 +1,56 @@
# frozen_string_literal: true
module Jobs
class PostsLocaleDetectionBackfill < ::Jobs::Scheduled
every 5.minutes
sidekiq_options retry: false
cluster_concurrency 1
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
return if SiteSetting.ai_translation_backfill_rate == 0
posts =
Post
.where(locale: nil)
.where(deleted_at: nil)
.where("posts.user_id > 0")
.where.not(raw: [nil, ""])
if SiteSetting.ai_translation_backfill_limit_to_public_content
public_categories = Category.where(read_restricted: false).pluck(:id)
posts =
posts
.joins(:topic)
.where(topics: { category_id: public_categories })
.where(topics: { archetype: "regular" })
end
if SiteSetting.ai_translation_backfill_max_age_days > 0
posts =
posts.where(
"posts.created_at > ?",
SiteSetting.ai_translation_backfill_max_age_days.days.ago,
)
end
posts = posts.order(updated_at: :desc).limit(SiteSetting.ai_translation_backfill_rate)
return if posts.empty?
posts.each do |post|
begin
DiscourseAi::Translation::PostLocaleDetector.detect_locale(post)
rescue FinalDestination::SSRFDetector::LookupFailedError
# do nothing, there are too many sporadic lookup failures
rescue => e
DiscourseAi::Translation::VerboseLogger.log(
"Failed to detect post #{post.id}'s locale: #{e.message}",
)
end
end
DiscourseAi::Translation::VerboseLogger.log("Detected #{posts.size} post locales")
end
end
end
@@ -0,0 +1,18 @@
# frozen_string_literal: true
module Jobs
class TopicLocalizationBackfill < ::Jobs::Scheduled
every 5.minutes
cluster_concurrency 1
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
return if SiteSetting.experimental_content_localization_supported_locales.blank?
return if SiteSetting.ai_translation_backfill_rate == 0
Jobs.enqueue(:localize_topics, limit: SiteSetting.ai_translation_backfill_rate)
end
end
end
@@ -0,0 +1,47 @@
# frozen_string_literal: true
module Jobs
class TopicsLocaleDetectionBackfill < ::Jobs::Scheduled
every 5.minutes
sidekiq_options retry: false
cluster_concurrency 1
def execute(args)
return if !SiteSetting.discourse_ai_enabled
return if !SiteSetting.ai_translation_enabled
limit = SiteSetting.ai_translation_backfill_rate
return if limit == 0
topics = Topic.where(locale: nil, deleted_at: nil).where("topics.user_id > 0")
if SiteSetting.ai_translation_backfill_limit_to_public_content
topics = topics.where(category_id: Category.where(read_restricted: false).select(:id))
end
if SiteSetting.ai_translation_backfill_max_age_days > 0
topics =
topics.where(
"topics.created_at > ?",
SiteSetting.ai_translation_backfill_max_age_days.days.ago,
)
end
topics = topics.order(updated_at: :desc).limit(limit)
return if topics.empty?
topics.each do |topic|
begin
DiscourseAi::Translation::TopicLocaleDetector.detect_locale(topic)
rescue FinalDestination::SSRFDetector::LookupFailedError
# do nothing, there are too many sporadic lookup failures
rescue => e
DiscourseAi::Translation::VerboseLogger.log(
"Failed to detect topic #{topic.id}'s locale: #{e.message}",
)
end
end
DiscourseAi::Translation::VerboseLogger.log("Detected #{topics.size} topic locales")
end
end
end
+1
View File
@@ -266,6 +266,7 @@ class AiPersona < ActiveRecord::Base
define_method(:top_p) { @ai_persona&.top_p }
define_method(:system_prompt) { @ai_persona&.system_prompt || "You are a helpful bot." }
define_method(:uploads) { @ai_persona&.uploads }
define_method(:response_format) { @ai_persona&.response_format }
define_method(:examples) { @ai_persona&.examples }
end
end
+27
View File
@@ -22,6 +22,8 @@ class AiTool < ActiveRecord::Base
message: I18n.t("discourse_ai.tools.name.characters"),
}
validate :validate_parameters_enum
def signature
{
name: function_call_name,
@@ -57,6 +59,30 @@ class AiTool < ActiveRecord::Base
end
end
def validate_parameters_enum
return unless parameters.is_a?(Array)
parameters.each_with_index do |param, index|
next if !param.is_a?(Hash) || !param.key?("enum")
enum_values = param["enum"]
if enum_values.empty?
errors.add(
:parameters,
"Parameter '#{param["name"]}' at index #{index}: enum cannot be empty",
)
next
end
if enum_values.uniq.length != enum_values.length
errors.add(
:parameters,
"Parameter '#{param["name"]}' at index #{index}: enum values must be unique",
)
end
end
end
def self.preamble
<<~JS
/**
@@ -142,6 +168,7 @@ class AiTool < ActiveRecord::Base
* base_64_content (string): Base64 encoded content of the file.
* Returns: { id: number, url: string, short_url: string } - Details of the created upload record.
*
* upload.getUrl(shortUrl): Given a short URL, eg upload://12345, returns the full CDN friendly URL of the upload.
* 5. chain
* Controls the execution flow.
*
+25
View File
@@ -0,0 +1,25 @@
# frozen_string_literal: true
class InferredConcept < ActiveRecord::Base
has_many :inferred_concept_topics
has_many :topics, through: :inferred_concept_topics
has_many :inferred_concept_posts
has_many :posts, through: :inferred_concept_posts
validates :name, presence: true, uniqueness: true
end
# == Schema Information
#
# Table name: inferred_concepts
#
# id :bigint not null, primary key
# name :string not null
# created_at :datetime not null
# updated_at :datetime not null
#
# Indexes
#
# index_inferred_concepts_on_name (name) UNIQUE
#
+25
View File
@@ -0,0 +1,25 @@
# frozen_string_literal: true
class InferredConceptPost < ActiveRecord::Base
belongs_to :inferred_concept
belongs_to :post
validates :inferred_concept_id, presence: true
validates :post_id, presence: true
validates :inferred_concept_id, uniqueness: { scope: :post_id }
end
# == Schema Information
#
# Table name: inferred_concept_posts
#
# inferred_concept_id :bigint
# post_id :bigint
# created_at :datetime not null
# updated_at :datetime not null
#
# Indexes
#
# index_inferred_concept_posts_on_inferred_concept_id (inferred_concept_id)
# index_inferred_concept_posts_uniqueness (post_id,inferred_concept_id) UNIQUE
#
+25
View File
@@ -0,0 +1,25 @@
# frozen_string_literal: true
class InferredConceptTopic < ActiveRecord::Base
belongs_to :inferred_concept
belongs_to :topic
validates :inferred_concept_id, presence: true
validates :topic_id, presence: true
validates :inferred_concept_id, uniqueness: { scope: :topic_id }
end
# == Schema Information
#
# Table name: inferred_concept_topics
#
# inferred_concept_id :bigint
# topic_id :bigint
# created_at :datetime not null
# updated_at :datetime not null
#
# Indexes
#
# index_inferred_concept_topics_on_inferred_concept_id (inferred_concept_id)
# index_inferred_concept_topics_uniqueness (topic_id,inferred_concept_id) UNIQUE
#
@@ -0,0 +1,34 @@
# frozen_string_literal: true
class AiInferredConceptPostSerializer < ApplicationSerializer
attributes :id,
:post_number,
:topic_id,
:topic_title,
:username,
:avatar_template,
:created_at,
:updated_at,
:excerpt,
:truncated,
:inferred_concepts
def avatar_template
User.avatar_template(object.username, object.uploaded_avatar_id)
end
def excerpt
Post.excerpt(object.cooked)
end
def truncated
object.cooked.length > SiteSetting.post_excerpt_maxlength
end
def inferred_concepts
ActiveModel::ArraySerializer.new(
object.inferred_concepts,
each_serializer: InferredConceptSerializer,
)
end
end
@@ -0,0 +1,5 @@
# frozen_string_literal: true
class InferredConceptSerializer < ApplicationSerializer
attributes :id, :name, :created_at, :updated_at
end
@@ -1,6 +1,7 @@
import Component from "@glimmer/component";
import { tracked } from "@glimmer/tracking";
import { action } from "@ember/object";
import didInsert from "@ember/render-modifiers/modifiers/did-insert";
import { service } from "@ember/service";
import DButton from "discourse/components/d-button";
import htmlClass from "discourse/helpers/html-class";
@@ -88,6 +89,15 @@ export default class AiArtifactComponent extends Component {
}`;
}
@action
setDataAttributes(element) {
if (this.args.dataAttributes) {
Object.entries(this.args.dataAttributes).forEach(([key, value]) => {
element.setAttribute(key, value);
});
}
}
<template>
{{#if this.expanded}}
{{htmlClass "ai-artifact-expanded"}}
@@ -118,6 +128,7 @@ export default class AiArtifactComponent extends Component {
src={{this.artifactUrl}}
width="100%"
frameborder="0"
{{didInsert this.setDataAttributes}}
></iframe>
{{/if}}
{{#unless this.requireClickToRun}}
File diff suppressed because it is too large Load Diff
@@ -10,6 +10,7 @@ import { composeAiBotMessage } from "../lib/ai-bot-helper";
import { AI_CONVERSATIONS_PANEL } from "../services/ai-conversations-sidebar-manager";
export default class AiBotHeaderIcon extends Component {
@service appEvents;
@service composer;
@service currentUser;
@service navigationMenu;
@@ -51,6 +52,7 @@ export default class AiBotHeaderIcon extends Component {
}
if (this.siteSettings.ai_bot_enable_dedicated_ux) {
this.appEvents.trigger("discourse-ai:bot-header-icon-clicked");
return this.router.transitionTo("discourse-ai-bot-conversations");
}
@@ -5,6 +5,7 @@ import DButton from "discourse/components/d-button";
import { AI_CONVERSATIONS_PANEL } from "../services/ai-conversations-sidebar-manager";
export default class AiBotSidebarNewConversation extends Component {
@service appEvents;
@service router;
@service sidebarState;
@@ -14,6 +15,8 @@ export default class AiBotSidebarNewConversation extends Component {
@action
routeTo() {
this.appEvents.trigger("discourse-ai:new-conversation-btn-clicked");
if (this.router.currentRouteName !== "discourse-ai-bot-conversations") {
this.router.transitionTo("/discourse-ai/ai-bot/conversations");
}
@@ -45,7 +45,7 @@ export default class AiPersonaCollapsableExample extends Component {
(if (eq pairIdx 0) "user" "model")
)
}}
@validation="required|length:1,100"
@validation="required|length:1,5000"
@disabled={{@system}}
as |field|
>

Some files were not shown because too many files have changed in this diff Show More