We were passing: `{"category": 4}` which was then escape by Ember, basically causing a double Json stringify and preventing code to work down the road.
### ✨ What's new?
- [X] show sentiment counts by color on doughnut without needing to hover
- [X] minor UI padding adjustments
- [X] hide filters that are not allowed to be adjusted in selected chart view
- [X] fix z-index issues with drill-down filters
Adds a new rake task to generate topics using the AI helper model.
Generate topics using
```
bundle exec rake "ai:generate_topics[Don Quijote de la Mancha\, Cien años de soledad\, Os Lusíadas\,三国演义]"
```
For now it randomly chooses a non-read-restricted category and a random bunch of users.
This PR enhances the LLM triage automation with several important improvements:
- Add ability to use AI personas for automated replies instead of canned replies
- Add support for whisper responses
- Refactor LLM persona reply functionality into a reusable method
- Add new settings to configure response behavior in automations
- Improve error handling and logging
- Fix handling of personal messages in the triage flow
- Add comprehensive test coverage for new features
- Make personas configurable with more flexible requirements
This allows for more dynamic and context-aware responses in automated workflows, with better control over visibility and attribution.
## LLM Persona Triage
- Allows automated responses to posts using AI personas
- Configurable to respond as regular posts or whispers
- Adds context-aware formatting for topics and private messages
- Provides special handling for topic metadata (title, category, tags)
## LLM Tool Triage
- Enables custom AI tools to process and respond to posts
- Tools can analyze post content and invoke personas when needed
- Zero-parameter tools can be used for automated workflows
- Not enabled in production yet
## Implementation Details
- Added new scriptable registration in discourse_automation/ directory
- Created core implementation in lib/automation/ modules
- Enhanced PromptMessagesBuilder with topic-style formatting
- Added helper methods for persona and tool selection in UI
- Extended AI Bot functionality to support whisper responses
- Added rate limiting to prevent abuse
## Other Changes
- Added comprehensive test coverage for both automation types
- Enhanced tool runner with LLM integration capabilities
- Improved error handling and logging
This feature allows forum admins to configure AI personas to automatically respond to posts based on custom criteria and leverage AI tools for more complex triage workflows.
Tool Triage has been disabled in production while we finalize details of new scripting capabilities.
**This PR includes a variety of updates to the Sentiment Analysis report:**
- [X] Conditionally showing sentiment reports based on `sentiment_enabled` setting
- [X] Sentiment reports should only be visible in sidebar if data is in the reports
- [X] Fix infinite loading of posts in drill down
- [x] Fix markdown emojis showing not showing as emoji representation
- [x] Drill down of posts should have URL
- [x] ~~Different doughnut sizing based on post count~~ [reverting and will address in follow-up (see: `/t/146786/47`)]
- [X] Hide non-functional export button
- [X] Sticky drill down filter nav
adds support for "thinking tokens" - a feature that exposes the model's reasoning process before providing the final response. Key improvements include:
- Add a new Thinking class to handle thinking content from LLMs
- Modify endpoints (Claude, AWS Bedrock) to handle thinking output
- Update AI bot to display thinking in collapsible details section
- Fix SEARCH/REPLACE blocks to support empty replacement strings and general improvements to artifact editing
- Allow configurable temperature in triage and report automations
- Various bug fixes and improvements to diff parsing
## 🔍 Overview
This update ensures that the streaming for discoveries is smoother, especially on first update.
## ➕ More details
To help with smoother streaming, the discovery preview (which was being tracked as a separate property in the JS logic) will be removed and the entire discovery content will be shown/hidden via the existing CSS. The preview was already receiving the full update even though it was visually hidden, so removing the separate property shouldn't have any negative performance hit. Visually hiding it with CSS only will help simplify the component and also allow for smoother streaming. We will instead remove the buffered streaming approach and instead use typing timers similar to what we did for streaming summarization.
No related tests as streaming animations are difficult to test.
This update improves some of the UI around sentiment analysis reports:
1. Improve titles so it is above and truncated when long
2. Change doughnut to only show total count
3. Ensures sentiment posts have dates
4. Ensure expand post doesn't appear on short posts
* FEATURE: full support for Sonnet 3.7
- Adds support for Sonnet 3.7 with reasoning on bedrock and anthropic
- Fixes regression where provider params were not populated
Note. reasoning tokens are hardcoded to minimum of 100 maximum of 65536
* FIX: open ai non reasoning models need to use deprecate max_tokens
* FIX: legacy reasoning models not working, missing provider params
1. Legacy reasoning models (o1-preview / o1-mini) do not support developer or system messages, do not use them.
2. LLM editor form not showing all provider params due to missing remap
* add system test
* FEATURE: Experimental search results from an AI Persona.
When a user searches discourse, we'll send the query to an AI Persona to provide additional context and enrich the results. The feature depends on the user being a member of a group to which the persona has access.
* Update assets/stylesheets/common/ai-blinking-animation.scss
Co-authored-by: Keegan George <[email protected]>
---------
Co-authored-by: Keegan George <[email protected]>
## 🔍 Overview
This update adds a new report page at `admin/reports/sentiment_analysis` where admins can see a sentiment analysis report for the forum grouped by either category or tags.
## ➕ More details
The report can breakdown either category or tags into positive/negative/neutral sentiments based on the grouping (category/tag). Clicking on the doughnut visualization will bring up a post list of all the posts that were involved in that classification with further sentiment classifications by post.
The report can additionally be sorted in alphabetical order or by size, as well as be filtered by either category/tag based on the grouping.
## 👨🏽💻 Technical Details
The new admin report is registered via the pluginAPi with `api.registerReportModeComponent` to register the custom sentiment doughnut report. However, when each doughnut visualization is clicked, a new endpoint found at: `/discourse-ai/sentiment/posts` is fetched to showcase posts classified by sentiments based on the respective params.
## 📸 Screenshots
The require for `discourse/config/environment` should point to the local user's core Discourse environment instead of the hardcoded path for Sam's env.
From [pgvector/pgvector](https://github.com/pgvector/pgvector) README
> With approximate indexes, filtering is applied after the index is scanned. If a condition matches 10% of rows, with HNSW and the default hnsw.ef_search of 40, only 4 rows will match on average. For more rows, increase hnsw.ef_search.
>
> Starting with 0.8.0, you can enable [iterative index scans](https://github.com/pgvector/pgvector#iterative-index-scans), which will automatically scan more of the index when needed.
Since we are stuck on 0.7.0 we are going the first option for now.
- Add non-contiguous search/replace support using ... syntax
- Add judge support for evaluating LLM outputs with ratings
- Improve error handling and reporting in eval runner
- Add full section replacement support without search blocks
- Add fabricators and specs for artifact diffing
- Track failed searches to improve debugging
- Add JS syntax validation for artifact versions in eval system
- Update prompt documentation with clear guidelines
* improve eval output
* move error handling
* llm as a judge
* fix spec
* small note on evals
* FEATURE: Native PDF support
This amends it so we use PDF Reader gem to extract text from PDFs
* This means that our simple pdf eval passes at last
* fix spec
* skip test in CI
* test file support
* Update lib/utils/image_to_text.rb
Co-authored-by: Alan Guo Xiang Tan <[email protected]>
* address pr comments
---------
Co-authored-by: Alan Guo Xiang Tan <[email protected]>
This PR introduces several enhancements and refactorings to the AI Persona and RAG (Retrieval-Augmented Generation) functionalities within the discourse-ai plugin. Here's a breakdown of the changes:
**1. LLM Model Association for RAG and Personas:**
- **New Database Columns:** Adds `rag_llm_model_id` to both `ai_personas` and `ai_tools` tables. This allows specifying a dedicated LLM for RAG indexing, separate from the persona's primary LLM. Adds `default_llm_id` and `question_consolidator_llm_id` to `ai_personas`.
- **Migration:** Includes a migration (`20250210032345_migrate_persona_to_llm_model_id.rb`) to populate the new `default_llm_id` and `question_consolidator_llm_id` columns in `ai_personas` based on the existing `default_llm` and `question_consolidator_llm` string columns, and a post migration to remove the latter.
- **Model Changes:** The `AiPersona` and `AiTool` models now `belong_to` an `LlmModel` via `rag_llm_model_id`. The `LlmModel.proxy` method now accepts an `LlmModel` instance instead of just an identifier. `AiPersona` now has `default_llm_id` and `question_consolidator_llm_id` attributes.
- **UI Updates:** The AI Persona and AI Tool editors in the admin panel now allow selecting an LLM for RAG indexing (if PDF/image support is enabled). The RAG options component displays an LLM selector.
- **Serialization:** The serializers (`AiCustomToolSerializer`, `AiCustomToolListSerializer`, `LocalizedAiPersonaSerializer`) have been updated to include the new `rag_llm_model_id`, `default_llm_id` and `question_consolidator_llm_id` attributes.
**2. PDF and Image Support for RAG:**
- **Site Setting:** Introduces a new hidden site setting, `ai_rag_pdf_images_enabled`, to control whether PDF and image files can be indexed for RAG. This defaults to `false`.
- **File Upload Validation:** The `RagDocumentFragmentsController` now checks the `ai_rag_pdf_images_enabled` setting and allows PDF, PNG, JPG, and JPEG files if enabled. Error handling is included for cases where PDF/image indexing is attempted with the setting disabled.
- **PDF Processing:** Adds a new utility class, `DiscourseAi::Utils::PdfToImages`, which uses ImageMagick (`magick`) to convert PDF pages into individual PNG images. A maximum PDF size and conversion timeout are enforced.
- **Image Processing:** A new utility class, `DiscourseAi::Utils::ImageToText`, is included to handle OCR for the images and PDFs.
- **RAG Digestion Job:** The `DigestRagUpload` job now handles PDF and image uploads. It uses `PdfToImages` and `ImageToText` to extract text and create document fragments.
- **UI Updates:** The RAG uploader component now accepts PDF and image file types if `ai_rag_pdf_images_enabled` is true. The UI text is adjusted to indicate supported file types.
**3. Refactoring and Improvements:**
- **LLM Enumeration:** The `DiscourseAi::Configuration::LlmEnumerator` now provides a `values_for_serialization` method, which returns a simplified array of LLM data (id, name, vision_enabled) suitable for use in serializers. This avoids exposing unnecessary details to the frontend.
- **AI Helper:** The `AiHelper::Assistant` now takes optional `helper_llm` and `image_caption_llm` parameters in its constructor, allowing for greater flexibility.
- **Bot and Persona Updates:** Several updates were made across the codebase, changing the string based association to a LLM to the new model based.
- **Audit Logs:** The `DiscourseAi::Completions::Endpoints::Base` now formats raw request payloads as pretty JSON for easier auditing.
- **Eval Script:** An evaluation script is included.
**4. Testing:**
- The PR introduces a new eval system for LLMs, this allows us to test how functionality works across various LLM providers. This lives in `/evals`
Prior to this commit, editing the provider wouldn't recompute the provider params. It would also not correctly recompute the "canEditURL" property.
To make possible this commit has:
- made a fix in core: https://github.com/discourse/discourse/pull/31329
- ensures the provider params are recomputed when provider is changed
- made the check on `canEditURL` based on form state and not initial model value
Tests have been added to confirm the expected behavior.
When you already have embeddings for a model stored and change models,
our backfill script was failing to backfill the newly configured model.
Regression introduced most likely in 1686a8a
Currently in core re-flagging something that is already flagged as spam
is not supported, long term we may want to support this but in the meantime
we should not be silencing/hiding if the PostActionCreator fails
when flagging things as spam.
---------
Co-authored-by: Ted Johansson <[email protected]>
* FEATURE: Tool name validation
- Add unique index to the name column of the ai_tools table
- correct our tests for AiToolController
- tool_name field which will be used to represent to LLM
- Add tool_name to Tools's presets
- Add duplicate tools validation for AiPersona
- Add unique constraint to the name column of the ai_tools table
* DEV: Validate duplicate tool_name between builin tools and custom tools
* lint
* chore: fix linting
* fix conlict mistakes
* chore: correct icon class
* chore: fix failed specs
* Add max_length to tool_name
* chore: correct the option name
* lintings
* fix lintings
Before this change, a summary was only outdated when new content appeared, for topics with "best replies", when the query returned different results. The intent behind this change is to detect when a summary is outdated as a result of an edit.
Additionally, we are changing the backfill candidates query to compare "ai_summary_backfill_topic_max_age_days" against "last_posted_at" instead of "created_at", to catch long-lived, active topics. This was discussed here: https://meta.discourse.org/t/ai-summarization-backfill-is-stuck-keeps-regenerating-the-same-topic/347088/14?u=roman_rizzi
### Why
This pull request fundamentally restructures how AI bots create and update web artifacts to address critical limitations in the previous approach:
1. **Improved Artifact Context for LLMs**: Previously, artifact creation and update tools included the *entire* artifact source code directly in the tool arguments. This overloaded the Language Model (LLM) with raw code, making it difficult for the LLM to maintain a clear understanding of the artifact's current state when applying changes. The LLM would struggle to differentiate between the base artifact and the requested modifications, leading to confusion and less effective updates.
2. **Reduced Token Usage and History Bloat**: Including the full artifact source code in every tool interaction was extremely token-inefficient. As conversations progressed, this redundant code in the history consumed a significant number of tokens unnecessarily. This not only increased costs but also diluted the context for the LLM with less relevant historical information.
3. **Enabling Updates for Large Artifacts**: The lack of a practical diff or targeted update mechanism made it nearly impossible to efficiently update larger web artifacts. Sending the entire source code for every minor change was both computationally expensive and prone to errors, effectively blocking the use of AI bots for meaningful modifications of complex artifacts.
**This pull request addresses these core issues by**:
* Introducing methods for the AI bot to explicitly *read* and understand the current state of an artifact.
* Implementing efficient update strategies that send *targeted* changes rather than the entire artifact source code.
* Providing options to control the level of artifact context included in LLM prompts, optimizing token usage.
### What
The main changes implemented in this PR to resolve the above issues are:
1. **`Read Artifact` Tool for Contextual Awareness**:
- A new `read_artifact` tool is introduced, enabling AI bots to fetch and process the current content of a web artifact from a given URL (local or external).
- This provides the LLM with a clear and up-to-date representation of the artifact's HTML, CSS, and JavaScript, improving its understanding of the base to be modified.
- By cloning local artifacts, it allows the bot to work with a fresh copy, further enhancing context and control.
2. **Refactored `Update Artifact` Tool with Efficient Strategies**:
- The `update_artifact` tool is redesigned to employ more efficient update strategies, minimizing token usage and improving update precision:
- **`diff` strategy**: Utilizes a search-and-replace diff algorithm to apply only the necessary, targeted changes to the artifact's code. This significantly reduces the amount of code sent to the LLM and focuses its attention on the specific modifications.
- **`full` strategy**: Provides the option to replace the entire content sections (HTML, CSS, JavaScript) when a complete rewrite is required.
- Tool options enhance the control over the update process:
- `editor_llm`: Allows selection of a specific LLM for artifact updates, potentially optimizing for code editing tasks.
- `update_algorithm`: Enables choosing between `diff` and `full` update strategies based on the nature of the required changes.
- `do_not_echo_artifact`: Defaults to true, and by *not* echoing the artifact in prompts, it further reduces token consumption in scenarios where the LLM might not need the full artifact context for every update step (though effectiveness might be slightly reduced in certain update scenarios).
3. **System and General Persona Tool Option Visibility and Customization**:
- Tool options, including those for system personas, are made visible and editable in the admin UI. This allows administrators to fine-tune the behavior of all personas and their tools, including setting specific LLMs or update algorithms. This was previously limited or hidden for system personas.
4. **Centralized and Improved Content Security Policy (CSP) Management**:
- The CSP for AI artifacts is consolidated and made more maintainable through the `ALLOWED_CDN_SOURCES` constant. This improves code organization and future updates to the allowed CDN list, while maintaining the existing security posture.
5. **Codebase Improvements**:
- Refactoring of diff utilities, introduction of strategy classes, enhanced error handling, new locales, and comprehensive testing all contribute to a more robust, efficient, and maintainable artifact management system.
By addressing the issues of LLM context confusion, token inefficiency, and the limitations of updating large artifacts, this pull request significantly improves the practicality and effectiveness of AI bots in managing web artifacts within Discourse.
* DEV: raise timeout for reasoning LLMs
* FIX: use id to identify llms, not model_name
model_name is not unique, in the case of reasoning models
you may configure the same llm multiple times using different
reasoning levels.
This update resolves a regression that was introduced in https://github.com/discourse/discourse-ai/pull/1036/files. Previously, only seeded models that were allowed could be configured for model settings. However, in our attempts to prevent unreachable LLM errors from not allowing settings to persist, it also unknowingly allowed seeded models that were not allowed to be configured. This update resolves this issue, while maintaining the ability to still set unreachable LLMs.
Before this change, we let you set the embeddings selected model back to " " even with embeddings enabled. This will leave the site in a broken state.
Additionally, it adds a fail-safe for these scenarios to avoid errors on the topics page.
We have removed this flag in core. All plugins now use the "top mode" for their navigation. A backwards-compatible change has been made in core while we remove the usage from plugins.
This change fixes two different problems.
First, we add a data migration to migrate the configuration of sites using Open AI's embedding model. There was a window between the embedding config changes and #1087, where sites could end up in a broken state due to an unconfigured selected model setting, as reported on https://meta.discourse.org/t/-/348964
The second fix drops pre-seeded search indexes of the models we didn't migrate and corrects the ones where the dimensions don't match. Since the index uses the model ID, new embedding configs could use one of these ones even when the dimensions no longer match.
This should give us a better idea on how our scanner is faring across sites.
```
# HELP discourse_discourse_ai_spam_detection AI spam scanning statistics
# TYPE discourse_discourse_ai_spam_detection counter
discourse_discourse_ai_spam_detection{db="default",type="scanned"} 16
discourse_discourse_ai_spam_detection{db="default",type="is_spam"} 7
discourse_discourse_ai_spam_detection{db="default",type="false_positive"} 1
discourse_discourse_ai_spam_detection{db="default",type="false_negative"} 2
```
We have a flag to signal we are shortening the embeddings of a model.
Only used in Open AI's text-embedding-3-*, but we plan to use it for other services.
* Use AR model for embeddings features
* endpoints
* Embeddings CRUD UI
* Add presets. Hide a couple more settings
* system specs
* Seed embedding definition from old settings
* Generate search bit index on the fly. cleanup orphaned data
* support for seeded models
* Fix run test for new embedding
* fix selected model not set correctly
This adds registration and last known IP information and email to scanning context.
This provides another hint for spam scanner about possible malicious users.
For example registered in India, replying from Australia or
email is clearly a throwaway email address.
To quickly select backfill candidates without comparing SHAs, we compare the last summarized post to the topic's highest_post_number. However, hiding or deleting a post and adding a small action will update this column, causing the job to stall and re-generate the same summary repeatedly until someone posts a regular reply. On top of this, this is not always true for topics with `best_replies`, as this last reply isn't necessarily included.
Since this is not evident at first glance and each summarization strategy picks its targets differently, I'm opting to simplify the backfill logic and how we track potential candidates.
The first step is dropping `content_range`, which serves no purpose and it's there because summary caching was supposed to work differently at the beginning. So instead, I'm replacing it with a column called `highest_target_number`, which tracks `highest_post_number` for topics and could track other things like channel's `message_count` in the future.
Now that we have this column when selecting every potential backfill candidate, we'll check if the summary is truly outdated by comparing the SHAs, and if it's not, we just update the column and move on
When enabling spam scanner it there may be old unscanned posts
this can create a risky situation where spam scanner operates
on legit posts during false positives
To keep this a lot safer we no longer try to hide old stuff by
the spammers.
This update fixes an issue when the composer helper menu was not being shown on tablets in desktop mode. Updating the `z-index` to use the modal-dialog case is more appropriate here.
Adds a comprehensive quota management system for LLM models that allows:
- Setting per-group (applied per user in the group) token and usage limits with configurable durations
- Tracking and enforcing token/usage limits across user groups
- Quota reset periods (hourly, daily, weekly, or custom)
- Admin UI for managing quotas with real-time updates
This system provides granular control over LLM API usage by allowing admins
to define limits on both total tokens and number of requests per group.
Supports multiple concurrent quotas per model and automatically handles
quota resets.
Co-authored-by: Keegan George <[email protected]>
Disabling streaming is required for models such o1 that do not have streaming
enabled yet
It is good to carry this feature around in case various apis decide not to support streaming endpoints and Discourse AI can continue to work just as it did before.
Also: fixes issue where sharing artifacts would miss viewport leading to tiny artifacts on mobile
This update adds some structure for handling errors in the spam config while also handling a specific error related to the spam scanning user not being an admin account.
The seeded LLM setting: `SiteSetting.ai_spam_detection_model_allowed_seeded_models` returns a _string_ with IDs separated by pipes. running `_map` on it will return an array with strings. We were previously checking for the id with custom prefix identifier, but instead we should be checking the stringified ID.
* FEATURE: smart date support for AI helper
This feature allows conversion of human typed in dates and times
to smart "Discourse" timezone friendly dates.
* fix specs and lint
* lint
* address feedback
* add specs
This PR fixes an issue where LLM enumerator would error out when `SiteSetting.ai_spam_detection = true` but there was no `AiModerationSetting.spam` present.
Typically, we add an `LlmDependencyValidator` for the setting itself, however, since Spam is unique in that it has it's model set in `AiModerationSetting` instead of a `SiteSetting`, we'll add a simple check here to prevent erroring out.
Previously we had the behaviour for model settings so that when you try and set a model, it runs a test and returns an error if it can't run the test successfully. The error then prevents you from setting the site setting.
This results in some issues when we try and automate things. This PR updates that so that the test runs and discreetly logs the changes, but doesn't prevent the setting from being set. Instead we rely on "run test" in the LLM config along with ProblemChecks to catch issues.
Previous version was prone to the bug:
https://github.com/ruby-concurrency/concurrent-ruby/issues/1075
This is particularly bad cause we could have a DB connection
attached to the thread and we never clear it up, so after N hours
this could start exhibiting weird connection issues.
- Add spam_score_type to AiSpamSerializer for better integration with reviewables.
- Introduce a custom filter for detecting AI spam false negatives in moderation workflows.
- Refactor spam report generation to improve identification of false negatives.
- Add tests to verify the custom filter and its behavior.
- Introduce links for all spam counts in report
This feature adds a periodic problem check which periodically checks for issues with LLMs that are in use. Periodically, we will run a test to see if the in use LLMs are still operational. If it is not, the LLM with the problem is surfaced to the admin so they can easily go and update the configuration.
This commit adds an "unavailable" state for the AI semantic search toggle. Currently the AI toggle disappears when the sort by is anything but Relevance which makes the UI confusing for users looking for AI results. This should help!
In a previous refactor, we moved the responsibility of querying and storing embeddings into the `Schema` class. Now, it's time for embedding generation.
The motivation behind these changes is to isolate vector characteristics in simple objects to later replace them with a DB-backed version, similar to what we did with LLM configs.
* FIX: Make sure gists are atleast five minutes old before updating them
* Update app/jobs/regular/fast_track_topic_gist.rb
Co-authored-by: Keegan George <[email protected]>
---------
Co-authored-by: Keegan George <[email protected]>
* REFACTOR: A Simpler way of interacting with embeddings' tables.
This change adds a new abstraction called `Schema`, which acts as a repository that supports the same DB features `VectorRepresentation::Base` has, with the exception that removes the need to have duplicated methods per embeddings table.
It is also a bit more flexible when performing a similarity search because you can pass it a block that gives you access to the builder, allowing you to add multiple joins/where conditions.
In this PR, we added functionality to hide the admin header for edit/new actions - https://github.com/discourse/discourse/pull/30175
To make it work properly, we have to rename `show` to `edit` which is also a more accurate name.
This introduces a comprehensive spam detection system that uses LLM models
to automatically identify and flag potential spam posts. The system is
designed to be both powerful and configurable while preventing false positives.
Key Features:
* Automatically scans first 3 posts from new users (TL0/TL1)
* Creates dedicated AI flagging user to distinguish from system flags
* Tracks false positives/negatives for quality monitoring
* Supports custom instructions to fine-tune detection
* Includes test interface for trying detection on any post
Technical Implementation:
* New database tables:
- ai_spam_logs: Stores scan history and results
- ai_moderation_settings: Stores LLM config and custom instructions
* Rate limiting and safeguards:
- Minimum 10-minute delay between rescans
- Only scans significant edits (>10 char difference)
- Maximum 3 scans per post
- 24-hour maximum age for scannable posts
* Admin UI features:
- Real-time testing capabilities
- 7-day statistics dashboard
- Configurable LLM model selection
- Custom instruction support
Security and Performance:
* Respects trust levels - only scans TL0/TL1 users
* Skips private messages entirely
* Stops scanning users after 3 successful public posts
* Includes comprehensive test coverage
* Maintains audit log of all scan attempts
---------
Co-authored-by: Keegan George <[email protected]>
Co-authored-by: Martin Brennan <[email protected]>
* UX: Improve rough edges of AI usage page
* Ensure all text uses I18n
* Change from <button> usage to <DButton>
* Use <AdminConfigAreaCard> in place of custom card styles
* Format numbers nicely using our number format helper,
show full values on hover using title attr
* Ensure 0 is always shown for counters, instead of being blank
* FEATURE: Load usage data after page load
Use ConditionalLoadingSpinner to hide load of usage
data, this prevents us hanging on page load with a white
screen.
* UX: Split users table, and add empty placeholders and page subheader
* DEV: Test fix
Instead of a stacked chart showing a separate series for positive and negative, this PR introduces a simplification to the overall sentiment dashboard. It comprises the sentiment into a single series of the difference between `positive - negative` instead. This should allow for the data to be more easy to scan and look for trends
Instead of a stacked chart showing a separate series for positive and negative, this PR introduces a simplification to the overall sentiment dashboard. It comprises the sentiment into a single series of the difference between `positive - negative` instead. This should allow for the data to be more easy to scan and look for trends.
* FEATURE: first class support for OpenRouter
This new implementation supports picking quantization and provider pref
Also:
- Improve logging for summary generation
- Improve error message when contacting LLMs fails
* Better support for full screen artifacts on iPad
Support back button to close full screen
Refactor dialect selection and add Nova API support
Change dialect selection to use llm_model object instead of just provider name
Add support for Amazon Bedrock's Nova API with native tools
Implement Nova-specific message processing and formatting
Update specs for Nova and AWS Bedrock endpoints
Enhance AWS Bedrock support to handle Nova models
Fix Gemini beta API detection logic
Previously, when clicking add footnote on an explain suggestion it would replace the selected word by finding the first occurrence of the word. This results in issues when there are more than one occurrences of a word in a post. This is not trivial to solve, so this PR instead prevents incorrect text replacements by only allowing the replacement if it's unique. We use the same logic here that we use to determine if something can be fast edited.
In this PR we also update tests for post helper explain suggestions. For a while, we haven't had tests here due to streaming/timing issues, we've been skipping our system specs. In this PR, we add acceptance tests to handle this which gives us improved ability to publish message bus updates in the testing environment so that it can be better tested without issues.
* FEATURE: Backfill posts sentiment.
It adds a scheduled job to backfill posts' sentiment, similar to our existing rake task, but with two settings to control the batch size and posts' max-age.
* Make sure model_name order is consistent.
For a while now we have not been sending the examples to AI
helper, which can lead to inconsistent results.
Note: this also means that in non English we did not send
English results, so this may end up reducing performance
That said first thing we need to do is fix the regression.
This PR fixes an issue where the tag suggester for edit title topic area was suggesting tags that are already assigned on a post. It also updates the amount of suggested tags to 7 so that there is still a decent amount of tags suggested when tags are already assigned.
Add support for versioned artifacts with improved diff handling
* Add versioned artifacts support allowing artifacts to be updated and tracked
- New `ai_artifact_versions` table to store version history
- Support for updating artifacts through a new `UpdateArtifact` tool
- Add version-aware artifact rendering in posts
- Include change descriptions for version tracking
* Enhance artifact rendering and security
- Add support for module-type scripts and external JS dependencies
- Expand CSP to allow trusted CDN sources (unpkg, cdnjs, jsdelivr, googleapis)
- Improve JavaScript handling in artifacts
* Implement robust diff handling system (this is dormant but ready to use once LLMs catch up)
- Add new DiffUtils module for applying changes to artifacts
- Support for unified diff format with multiple hunks
- Intelligent handling of whitespace and line endings
- Comprehensive error handling for diff operations
* Update routes and UI components
- Add versioned artifact routes
- Update markdown processing for versioned artifacts
Also
- Tweaks summary prompt
- Improves upload support in custom tool to also provide urls
In the tag suggester menu we use `DButton` as a wrapper element and use the `discourseTag` helper to render the text inside the element. So visually there is text content inside the button. However, since `DButton` assumes that no `label`/`translatedLabel` inside an element means `.no-text` CSS style should be applied to the button's element, it was resulting in some incorrect styling being applied to this menu. This PR resolves that by programmatically adding the tag as a `translatedLabel` and then visually hiding it with CSS.
The description of the automatic captions dialog prompt included "Would you like to enable automatic auto captions on image uploads." I removed the second "auto" at the suggestion of translators.
- Added a new admin interface to track AI usage metrics, including tokens, features, and models.
- Introduced a new route `/admin/plugins/discourse-ai/ai-usage` and supporting API endpoint in `AiUsageController`.
- Implemented `AiUsageSerializer` for structuring AI usage data.
- Integrated CSS stylings for charts and tables under `stylesheets/modules/llms/common/usage.scss`.
- Enhanced backend with `AiApiAuditLog` model changes: added `cached_tokens` column (implemented with OpenAI for now) with relevant DB migration and indexing.
- Created `Report` module for efficient aggregation and filtering of AI usage metrics.
- Updated AI Bot title generation logic to log correctly to user vs bot
- Extended test coverage for the new tracking features, ensuring data consistency and access controls.
This change adds a simpler class for sentiment classification, replacing the soon-to-be removed `Classificator` hierarchy. Additionally, it adds a method for classifying concurrently, speeding up the backfill rake task.
This PR updates the logic for the location map so it permits only the desired prompts through to the composer/post menu. Anything else won't be shown by default.
This PR also adds relevant tests to prevent regression.
### 🔍 Overview
With the recent changes to allow DiscourseAi in the translator plugin, `detect_text_locale` was needed as a CompletionPrompt. However, it is leaking into composer/post helper menus. This PR ensures we don't not show it in those menus.
This commit applies further admin UI guidelines, now that they have been more
fleshed out in core, to the AI admin UI:
* Tools
* LLMs
* Personas
The changes include but are not limited to:
* Applying the table CSS classes, for desktop and mobile
* Adding a description and learn more link for each tab
* Adding an empty list placeholder with CTA using `AdminConfigAreaEmptyList`
* Replacing custom headings with `AdminPageSubheader`
We are adding a new method for generating and storing embeddings in bulk, which relies on `Concurrent::Promises::Future`. Generating an embedding consists of three steps:
Prepare text
HTTP call to retrieve the vector
Save to DB.
Each one is independently executed on whatever thread the pool gives us.
We are bringing a custom thread pool instead of the global executor since we want control over how many threads we spawn to limit concurrency. We also avoid firing thousands of HTTP requests when working with large batches.
This spec fails inconsistently with:
-fragment-n14
+You are a helpful Discourse assistant.
+You _understand_ and **generate** Discourse Markdown.
+You live in a Discourse Forum Message.
+
+You live in the forum with the URL: http://test.localhost
+The title of your site: test site title
+The description is: test site description
+The participants in this conversation are: joe, jane
+The date now is: 2024-11-25 20:23:02 UTC, much has changed since you were trained.
+
+You were trained on OLD data, lean on search to get up to date information about this forum
+When searching try to SIMPLIFY search terms
+Discourse search joins all terms with AND. Reduce and simplify terms to find more results.<guidance>
+The following texts will give you additional guidance for your response.
+We included them because we believe they are relevant to this conversation topic.
+
+Texts:
+
+fragment-n10
+fragment-n9
+fragment-n8
+fragment-n7
+fragment-n6
+fragment-n5
+fragment-n4
+fragment-n3
+fragment-n2
+fragment-n1
+</guidance>
* FEATURE: allow mentioning an LLM mid conversation to switch
This is a edgecase feature that allow you to start a conversation
in a PM with LLM1 and then use LLM2 to evaluation or continue
the conversation
* FEATURE: allow auto silencing of spam accounts
New rule can also allow for silencing an account automatically
This can prevent spammers from creating additional posts.
Two changes worth mentioning:
`#instance` returns a fully configured embedding endpoint ready to use.
All endpoints respond to the same method and have the same signature - `perform!(text)`
This makes it easier to reuse them when generating embeddings in bulk.
* PERF: Preload only gists when including summaries in topic list
* Add unique index on summaries and dedup existing records
* Make hot topics batch size setting hidden
The `topic_query_create_list_topics` modifier we append was always meant to avoid an N+1 situation when serializing gists. However, I tried to be too smart and only preload these, which resulted in some topics with *only* regular summaries getting removed from the list. This issue became apparent now we are adding gists to other lists besides hot.
Let's simplify the preloading, which still solves the N+1 issue, and let the serializer get the needed summary.
* FIX: automatically bust cache for share ai assets
CDNs can be configured to strip query params in Discourse
hosting. This is generally safe, but in this case we had
no way of busting the cache using the path.
New design properly caches and properly breaks busts the
cache if asset changes so we don't need to worry about versions
* one day I will set up conditional lint on save :)
* FEATURE: Make emotion /filter ordering match the dashboard table
This change makes the /filter endpoint use the same criteria we use
in the dashboard table for emotion, so it is not confusing for users.
It means that only posts made in the period with the emotion shall be
shown in the /filter, and the order is simply a count of posts that
match the emotion in the period.
It also uses a trick to extract the filter period, and apply it to
the CTE clause that calculates post emotion count on the period, making
it a bit more efficient. Downside is that /filter filters are evaluated
from left to right, so it will only get the speed-up if the emotion
order is last. As we do this on the dashboard table, it should cover
most uses of the ordering, kicking the need for materialized views
down the road.
* Remove zero score in filter
* add table tooltip
* lint
1. Keep source in a "details" block after rendered so it does
not overwhelm users
2. Ensure artifacts are never indexed by robots
3. Cache break our CSS that changed recently
* FIX: Misc fixes for sentiment in the admin dashboard
- Fixes missing filters for the main graph
- Fixes previous 30 days trend in emotion table
Also moves links to individual cells in emotion table, so admins can
drill down to the specific time period on their reports.
* lints
We use `includes` instead of `joins` because we want to eager-load summaries, avoiding an extra query when summarizing. However, Rails will complain unless you explicitly inform them you plan to use that inside a `WHERE` clause.
The topic query is used differently, and we can't assume the modifier will always receive an AR relation. Let's scope it to `Discourse#filters` instead of most lists.
It's important that artifacts are never given 'same origin' access to the forum domain, so that they cannot access cookies, or make authenticated HTTP requests. So even when visiting the URL directly, we need to wrap them in a sandboxed iframe.
This is a significant PR that introduces AI Artifacts functionality to the discourse-ai plugin along with several other improvements. Here are the key changes:
1. AI Artifacts System:
- Adds a new `AiArtifact` model and database migration
- Allows creation of web artifacts with HTML, CSS, and JavaScript content
- Introduces security settings (`strict`, `lax`, `disabled`) for controlling artifact execution
- Implements artifact rendering in iframes with sandbox protection
- New `CreateArtifact` tool for AI to generate interactive content
2. Tool System Improvements:
- Adds support for partial tool calls, allowing incremental updates during generation
- Better handling of tool call states and progress tracking
- Improved XML tool processing with CDATA support
- Fixes for tool parameter handling and duplicate invocations
3. LLM Provider Updates:
- Updates for Anthropic Claude models with correct token limits
- Adds support for native/XML tool modes in Gemini integration
- Adds new model configurations including Llama 3.1 models
- Improvements to streaming response handling
4. UI Enhancements:
- New artifact viewer component with expand/collapse functionality
- Security controls for artifact execution (click-to-run in strict mode)
- Improved dialog and response handling
- Better error management for tool execution
5. Security Improvements:
- Sandbox controls for artifact execution
- Public/private artifact sharing controls
- Security settings to control artifact behavior
- CSP and frame-options handling for artifacts
6. Technical Improvements:
- Better post streaming implementation
- Improved error handling in completions
- Better memory management for partial tool calls
- Enhanced testing coverage
7. Configuration:
- New site settings for artifact security
- Extended LLM model configurations
- Additional tool configuration options
This PR significantly enhances the plugin's capabilities for generating and displaying interactive content while maintaining security and providing flexible configuration options for administrators.
Implement streaming tool call implementation for Anthropic and Open AI.
When calling:
llm.generate(..., partial_tool_calls: true) do ...
Partials may contain ToolCall instances with partial: true, These tool calls are partially populated with json partially parsed.
So for example when performing a search you may get:
ToolCall(..., {search: "hello" })
ToolCall(..., {search: "hello world" })
The library used to parse json is:
https://github.com/dgraham/json-stream
We use a fork cause we need access to the internal buffer.
This prepares internals to perform partial tool calls, but does not implement it yet.
This PR fixes an issue where the AI search results were not being reset when you append your search to an existing query param (typically when you've come from quick search). This is because `handleSearch()` doesn't get called in this situation. So here we explicitly check for query params, trigger a reset and search for those occasions.
This re-implements tool support in DiscourseAi::Completions::Llm #generate
Previously tool support was always returned via XML and it would be the responsibility of the caller to parse XML
New implementation has the endpoints return ToolCall objects.
Additionally this simplifies the Llm endpoint interface and gives it more clarity. Llms must implement
decode, decode_chunk (for streaming)
It is the implementers responsibility to figure out how to decode chunks, base no longer implements. To make this easy we ship a flexible json decoder which is easy to wire up.
Also (new)
Better debugging for PMs, we now have a next / previous button to see all the Llm messages associated with a PM
Token accounting is fixed for vllm (we were not correctly counting tokens)
This PR fixes an issue where clicking to regenerate a summary was still showing the cached summary. To resolve this we call resetSummary() to reset all the summarization related properties before creating a new request.
This PR further decouples the streaming animation by completely handling the streaming animation directly in the `AiSummaryBox` component. Previously, handling the streaming animation by calling methods in the `ai-streamer` API was leading to timing issues making things out-of-sync. This results in some issues such as the last update of streamed text not being shown. Handling streaming directly in the component should simplify things drastically and prevent any issues.
This change introduces a job to summarize topics and cache the results automatically. We provide a setting to control how many topics we'll backfill per hour and what the topic's minimum word count is to qualify.
We'll prioritize topics without summary over outdated ones.
Fixes encoding of params on LLM function calls.
Previously we would improperly return results if a function parameter returned an HTML tag.
Additionally adds some missing HTTP verbs to tool calls.
The custom field "discourse_ai_bypass_ai_reply" was added so
we can signal the post created hook to bypass replying even
if it thinks it should.
Otherwise there are cases where we double answer user questions
leading to much confusion.
This also slightly refactors code making the controller smaller
The new `/admin/plugins/discourse-ai/ai-personas/stream-reply.json` was added.
This endpoint streams data direct from a persona and can be used
to access a persona from remote systems leaving a paper trail in
PMs about the conversation that happened
This endpoint is only accessible to admins.
---------
Co-authored-by: Gabriel Grubba <[email protected]>
Co-authored-by: Keegan George <[email protected]>
The primary key is usually a bigint column, but the foreign key columns
are usually of integer type. This can lead to issues when joining these
columns due to mismatched types and different value ranges.
This was using a temporary plugin / test API to make tests pass, but it
is safe to alter "ai_document_fragment_embeddings" and
"rag_document_fragments" tables because they usually have less than 1M
rows and migration is going to be fast.
Depending on the size of the community, "classification_results" table
may have more than 1M rows and the migration will lock the table for a
longer time. However, classification runs in background jobs and they
will be automatically retried if they fail due to the lock, which makes
it acceptable.
* FEATURE: Fast-track gist regeneration when a hot topic gets a new post
* DEV: Introduce an upsert-like summarize
* FIX: Only enqueue fast-track gist for hot hot hot topics
---------
Co-authored-by: Rafael Silva <[email protected]>
* FIX/REFACTOR: FoldContent revamp
We hit a snag with our hot topic gist strategy: the regex we used to split the content didn't work, so we cannot send the original post separately. This was important for letting the model focus on what's new in the topic.
The algorithm doesn’t give us full control over how prompts are written, and figuring out how to format the content isn't straightforward. This means we're having to use more complicated workarounds, like regex.
To tackle this, I'm suggesting we simplify the approach a bit. Let's focus on summarizing as much as we can upfront, then gradually add new content until there's nothing left to summarize.
Also, the "extend" part is mostly for models with small context windows, which shouldn't pose a problem 99% of the time with the content volume we're dealing with.
* Fix fold docs
* Use #shift instead of #pop to get the first elem, not the last
This changeset contains 4 fixes:
1. We were allowing running tests on unsaved tools,
this is problematic cause uploads are not yet associated or indexed
leading to confusing results. We now only show the test button when
tool is saved.
2. We were not properly scoping rag document fragements, this
meant that personas and ai tools could get results from other
unrelated tools, just to be filtered out later
3. index.search showed options as "optional" but implementation
required the second option
4. When testing tools searching through document fragments was
not working at all cause we did not properly load the tool
* FIX: Llm selector / forced tools / search tool
This fixes a few issues:
1. When search was not finding any semantic results we would break the tool
2. Gemin / Anthropic models did not implement forced tools previously despite it being an API option
3. Mechanics around displaying llm selector were not right. If you disabled LLM selector server side persona PM did not work correctly.
4. Disabling native tools for anthropic model moved out of a site setting. This deliberately does not migrate cause this feature is really rare to need now, people who had it set probably did not need it.
5. Updates anthropic model names to latest release
* linting
* fix a couple of tests I missed
* clean up conditional
AI bot won't be turned on for seeded LLMs so it makes no sense to expose it here. This will cleanup the template and avoid the double `{{#unless}}` check.
A new feature_context json column was added to ai_api_audit_logs
This allows us to store rich json like context on any LLM request
made.
This new field now stores automation id and name.
Additionally allows llm_triage to specify maximum number of tokens
This means that you can limit the cost of llm triage by scanning only
first N tokens of a post.
In preparation for applying the streaming animation elsewhere, we want to better improve the organization of folder structure and methods used in the `ai-streamer`
This changeset:
1. Corrects some issues with "force_default_llm" not applying
2. Expands the LLM list page to show LLM usage
3. Clarifies better what "enabling a bot" on an llm means (you get it in the selector)
Previously, when we added smooth streaming animation to summarization (https://github.com/discourse/discourse-ai/pull/778) we used the same logic and lib we did for AI Bot. However, since `AiSummaryBox` is an Ember component, the direct DOM manipulation done in the streamer (`SummaryUpdater`) would often result in issues with summarization where sometimes summarization updates would hang, especially on the last result. This is likely due to the DOM manipulation being done in the streamer being incongruent with Ember's way of rendering.
In this PR, we remove the direct DOM manipulation done in the lib `SummaryUpdater` in favour of directly updating the properties in `AiSummaryBox` using the `componentContext`. Instead of messing with Ember's rendered DOM, passing the updates and allowing the component to render the updates directly should likely prevent further issues with summarization.
The bug itself is quite difficult to repro and also difficult to test, so no tests have been added to this PR. But I will be manually testing and assessing for any potential issues.
* Display gists in the hot topics list
* Adjust hot topics gist strategy and add a job to generate gists
* Replace setting with a configurable batch size
* Avoid loading summaries for other topic lists
* Tweak gist prompt to focus on latest posts in the context of the OP
* Remove serializer hack and rely on core change from discourse/discourse#29291
* Update lib/summarization/strategies/hot_topic_gists.rb
Co-authored-by: Rafael dos Santos Silva <[email protected]>
---------
Co-authored-by: Rafael dos Santos Silva <[email protected]>
Splits persona permissions so you can allow a persona on:
- chat dms
- personal messages
- topic mentions
- chat channels
(any combination is allowed)
Previously we did not have this flexibility.
Additionally, adds the ability to "tether" a language model to a persona so it will always be used by the persona. This allows people to use a cheaper language model for one group of people and more expensive one for other people
On very large sites, the rare cache misses for Related Topics can take around 200ms, which affects our p99 metric on the topic page. In order to mitigate this impact, we now have several tools at our disposal.
First, one is to migrate the index embedding type from halfvec to bit and change the related topic query to leverage the new bit index by changing the search algorithm from inner product to Hamming distance. This will reduce our index sizes by 90%, severely reducing the impact of embeddings on our storage. By making the related query a bit smarter, we can have zero impact on recall by using the index to over-capture N*2 results, then re-ordering those N*2 using the full halfvec vectors and taking the top N. The expected impact is to go from 200ms to <20ms for cache misses and from a 2.5GB index to a 250MB index on a large site.
Another tool is migrating our index type from IVFFLAT to HNSW, which can increase the cache misses performance even further, eventually putting us in the under 5ms territory.
Co-authored-by: Roman Rizzi <[email protected]>
This introduces another configuration that allows operators to
limit the amount of interactions with forced tool usage.
Forced tools are very handy in initial llm interactions, but as
conversation progresses they can hinder by slowing down stuff
and adding confusion.
The primary key is usually a bigint column, but the foreign key columns
usually are of integer type. This can lead to issues when joining these
columns due to mismatched types and different value ranges.
In a recent core change, all bigint sequences will start at a very high
value in the test environment to surface this type of errors. The same
change also added a temporary API that changes the column type to bigint
in order to allow for the tests to run.
The plugin API is only temporary and it is important for these plugins
to migrate their columns to bigint to avoid issues in the future.
This adds chain halting (ability to terminate llm chain in a tool)
and the ability to create uploads in a tool
Together this lets us integrate custom image generators into a
custom tool.
* FEATURE: allows forced LLM tool use
Sometimes we need to force LLMs to use tools, for example in RAG
like use cases we may want to force an unconditional search.
The new framework allows you backend to force tool usage.
Front end commit to follow
* UI for forcing tools now works, but it does not react right
* fix bugs
* fix tests, this is now ready for review
Previous to this change we could flag, but there was no way
to hide content and treat the flag as spam.
We had the option to hide topics, but this is not desirable for
a spam reply.
New option allows triage to hide a post if it is a reply, if the
post happens to be the first post on the topic, the topic will
be hidden.
This PR updates the rate limits for AI helper so that image caption follows a specific rate limit of 20 requests per minute. This should help when uploading multiple files that need to be captioned. This PR also updates the UI so that it shows toast message with the extracted error message instead of having a blocking `popupAjaxError` error dialog.
---------
Co-authored-by: Rafael dos Santos Silva <[email protected]>
Co-authored-by: Penar Musaraj <[email protected]>
Relies on https://github.com/discourse/discourse/pull/28477,
uses AdminSectionLandingWrapper and AdminSectionLandingItem
for the section items on the LLM page which are used to create
a new LLM config from a template.
This allows our users to add the Ollama provider and use it to serve our AI bot (completion/dialect).
In this PR, we introduce:
DiscourseAi::Completions::Dialects::Ollama which would help us translate by utilizing Completions::Endpoint::Ollama
Correct extract_completion_from and partials_from in Endpoints::Ollama
Also
Add tests for Endpoints::Ollama
Introduce ollama_model fabricator
This allows custom tools access to uploads and sophisticated searches using embedding.
It introduces:
- A shared front end for listing and uploading files (shared with personas)
- Backend implementation of index.search function within a custom tool.
Custom tools now may search through uploaded files
function invoke(params) {
return index.search(params.query)
}
This means that RAG implementers now may preload tools with knowledge and have high fidelity over
the search.
The search function support
specifying max results
specifying a subset of files to search (from uploads)
Also
- Improved documentation for tools (when creating a tool a preamble explains all the functionality)
- uploads were a bit finicky, fixed an edge case where the UI would not show them as updated
Restructures LLM config page so it is far clearer.
Also corrects bugs around adding LLMs and having LLMs not editable post addition
---------
Co-authored-by: Sam Saffron <[email protected]>
Often it is helpful to have the summary box open while composing a reply to the topic. However, the summary box currently gets closed each time you click outside the box. In this PR we add `closeOnClickOutside: false` attribute to the `DMenu` options for summary box to prevent that from occurring.
Previously we had some hardcoded markup with scss making a loading indicator wave. This code was being duplicated and used in both semantic search and summarization. We want to add the indicator wave to the AI helper diff modal as well and have the text flashing instead of the loading spinner. To ensure we do not repeat ourselves, in this PR we turn the summary indicator wave into a reusable template only component called: `AiIndicatorWave`. We then apply the usage of that component to semantic search, summarization, and the composer helper modal.
This commit fixes an issue where the composer AI helper was not visible on iPad in DiscourseHub. This was due to the z-index being different for `reply-control` when Discourse Hub inserts its `footer-nav`
Caveats
- No streaming, by design
- No tool support (including no XML tools)
- No vision
Open AI will revamt the model and more of these features may
become available.
This solution is a bit hacky for now
The `DiffModal` is triggered after selecting an option in the composer helper menu. After selecting an option, we should close the composer helper menu and only show the diff modal. On mobile, there was an edge-case where `this.args.close()` for was causing the closing of both the `DiffModal` and the `AiComposerHelperMenu`. This PR resolves that by ensuring the menu is closed _first_ asynchronously, followed by opening the relevant modal.
Polymorphic RAG means that we will be able to access RAG fragments both from AiPersona and AiCustomTool
In turn this gives us support for richer RAG implementations.
Previously we had moved the AI helper from the options menu to a selection menu that appears when selecting text in the composer. This had the benefit of making the AI helper a more discoverable feature. Now that some time has passed and the AI helper is more recognized, we will be moving it back to the composer toolbar.
This is better because:
- It consistent with other behavior and ways of accessing tools in the composer
- It has an improved mobile experience
- It reduces unnecessary code and keeps things easier to migrate when we have composer V2.
- It allows for easily triggering AI helper for all content by clicking the button instead of having to select everything.
Embedding search is rate limited due to potentially expensive
hyde operation (which require LLM access).
Embedding generally is very cheap compared to it. (usually 100x cheaper)
This raises the limit to 100 per minute for embedding searches,
while keeping the old 4 per minute for HyDE powered search.
Previously we waited 1 minute before automatically titling PMs
The new change introduces adding a title immediately after the the
llm replies
Prompt was also modified to include the LLM reply in title suggestion.
This helps situation like:
user: tell me a joke
llm: a very funy joke about horses
Then the title would be "A Funny Horse Joke"
Specs already covered some auto title logic, amended to also
catch the new message bus message we have been sending.
* FIX: we were never reindexing old content
Embedding backfill contains logic for searching for old content
change and then backfilling.
Unfortunately it was excluding all topics that had embedding
unconditionally, leading to no backfill ever happening.
This change adds a test and ensures we backfill.
* over select results, this ensures we will be more likely to find
ai results when filtered
This improves the site setting search so it performs a somewhat
fuzzy match.
Previously it did not handle seperators such as "space" and a
term such as "min_post_length" would not find "min_first_post_length"
A more liberal search algorithm makes it easier to the AI to
navigate settings.
* Minor fix, {{and parameter.enum parameter.enum.length}} is non
obviously broken.
If parameter.enum is a tracked array it will return the object
cause embers and helper implementation.
This corrects an issue where enum keeps on selecting itself by
mistake.
This allows callers of embedding based search to bypass hyde.
Hyde will expand the search term using an LLM, but if an LLM is
performing the search we can skip this expansion.
It also introduced some tests for the controller which we did not have
Previously there was too much work proofreading text, new implementation
provides a single shortcut and easy way of proofreading text.
Co-authored-by: Martin Brennan <[email protected]>
`withEventValue` is not needed here, because the `onChange` event comes from ace, not a normal DOM event. But even with that fix, it seems AceEditor doesn't yet work well with the DDAU pattern. On every keypress, the editor re-renders and puts the cursor back at the beginning.
For now, this commit removes the `@onChange` hook, so we go back to relying on the two-way binding of `@content`.
Followup to a5a39dd2ee
* FEATURE: LLM Triage support for systemless models.
This change adds support for OSS models without support for system messages. LlmTriage's system message field is no longer mandatory. We now send the post contents in a separate user message.
* Models using Ollama can also disable system prompts
New `ai_pm_summarization_allowed_groups` can be used to allow
visibility of the summarization feature on PMs.
This can be useful on forums where a lot of communication happens
inside PMs.
When navigating between topic we were not correctly resetting
internal state for summarization. This leads to a situation where
incorrect summaries can be displayed to users and wrong summaries
can be displayed.
Additionally our controller for grabbing summaries was always
streaming results via message bus, which could be delayed when
sidekiq is overloaded. We now will return the cached summary
right away if it is available direct from REST endpoint.
Creating a new model, either manually or from presets, doesn't initialize the `provider_params` object, meaning their custom params won't persist.
Additionally, this change adds some validations for Bedrock params, which are mandatory, and a clear message when a completion fails because we cannot build the URL.
- Validate fields to reduce the chance of breaking features by a misconfigured model.
- Fixed a bug where the URL might get deleted during an update.
- Display a warning when a model is currently in use.
* FIX: Add tool support to open ai compatible dialect and vllm
Automatic tools are in progress in vllm see: https://github.com/vllm-project/vllm/pull/5649
Even when they are supported, initial support will be uneven, only some models have native tool support
notably mistral which has some special tokens for tool support.
After the above PR lands in vllm we will still need to swap to
XML based tools on models without native tool support.
* fix specs
* DEV: Remove old code now that features rely on LlmModels.
* Hide old settings and migrate persona llm overrides
* Remove shadowing special URL + seeding code. Use srv:// prefix instead.
* FIX: Properly fix ai_summaries table sequence
Previous attempt at 3815360 could fail due to a race introduced in 1b0ba91 where summaries are migrated to core in a post_migrate erroneously.
Using RAG fragments can lead to considerably big system messages, which becomes problematic when models have a smaller context window.
Before this change, we only look at the rest of the conversation to make sure we don't surpass the limit, which could lead to two unwanted scenarios when having large system messages:
All other messages are excluded due to size.
The system message already exceeds the limit.
As a result, I'm putting a hard-limit of 60% of available tokens. We don't want to aggresively truncate because if rag fragments are included, the system message contains a lot of context to improve the model response, but we also want to make room for the recent messages in the conversation.
Followup 8d4a67fbe2
The prompt worked better, but it took the instructions
about never using lowercase a little too literally,
it wasn't using it for things like LLM or Discourse,
also it was almost always framing the title as questions
so now I asked it for a mix of questions and statements
because that's less ambiguous.
Changes the title generator prompt to avoid clickbait-y
titles and also try to avoid AI's favourite title format,
which is "Some Thing: Other Thing"
Leaving the chat thread title generator for now, that's
not as important, the bizarre titles add to the experience there.
* Seeding the SRV-backed model should happen inside an initializer.
* Keep the model up to date when the hidden setting changes.
* Use the correct Mixtral model name and fix previous data migration.
* URL validation should trigger only when we attempt to update it.
Using assistant role for system produces an error because
they expect alternating roles like user/assistant/user and so on.
Prompts cannot start with the assistant role.
1. Repairs the identity on the summary table, we migrated data without resetting it.
2. Adds an index into ai_summary table to match expected retrieval pattern
This allows summary to use the new LLM models and migrates of API key based model selection
Claude 3.5 etc... all work now.
---------
Co-authored-by: Roman Rizzi <[email protected]>
Follow up to b863ddc94b
Ruby:
* Validate `summary` (the column is `not null`)
* Fix `name` validation (the column has `max_length` 100)
* Fix table annotations
* Accept missing `parameter` attributes (`required, `enum`, `enum_values`)
JS:
* Use native classes
* Don't use ember's array extensions
* Add explicit service injections
* Correct class names
* Use `||=` operator
* Use `store` service to create records
* Remove unused service injections
* Extract consts
* Group actions together
* Use `async`/`await`
* Use `withEventValue`
* Sort html attributes
* Use DButtons `@label` arg
* Use `input` elements instead of Ember's `Input` component (same w/ textarea)
* Remove `btn-default` class (automatically applied by DButton)
* Don't mix `I18n.t` and `i18n` in the same template
* Don't track props that aren't used in a template
* Correct invalid `target.value` code
* Remove unused/invalid `this.parameter`/`onChange` code
* Whitespace
* Use the new service import `inject as service` -> `service`
* Use `Object.entries()`
* Add missing i18n strings
* Fix an error in `addEnumValue` (calling `pushObject` on `undefined`)
* Use `TrackedArray`/`TrackedObject`
* Transform tool `parameters` keys (`enumValues` -> `enum_values`)
* FIX: Use base64 encoded images in AI Image Caption via LLaVa
This fixed a regression introduced in #646 where we started sending
schemaless URLs for our LLaVa service, which doesn't handle it well.
Moving to base64 encoded images solves:
- The service needing to download images
Now the service running LLaVa doesn't need internet access
- Secure uploads compat
Every image is treated the same, less branching for secure uploads
- Image Size problems
Discourse is now responsible for ensure a max size for images
- Troublesome dev env
Previously to this commit you would need a dev env that was internet
acessible to use llava image captions
Introduces custom AI tools functionality.
1. Why it was added:
The PR adds the ability to create, manage, and use custom AI tools within the Discourse AI system. This feature allows for more flexibility and extensibility in the AI capabilities of the platform.
2. What it does:
- Introduces a new `AiTool` model for storing custom AI tools
- Adds CRUD (Create, Read, Update, Delete) operations for AI tools
- Implements a tool runner system for executing custom tool scripts
- Integrates custom tools with existing AI personas
- Provides a user interface for managing custom tools in the admin panel
3. Possible use cases:
- Creating custom tools for specific tasks or integrations (stock quotes, currency conversion etc...)
- Allowing administrators to add new functionalities to AI assistants without modifying core code
- Implementing domain-specific tools for particular communities or industries
4. Code structure:
The PR introduces several new files and modifies existing ones:
a. Models:
- `app/models/ai_tool.rb`: Defines the AiTool model
- `app/serializers/ai_custom_tool_serializer.rb`: Serializer for AI tools
b. Controllers:
- `app/controllers/discourse_ai/admin/ai_tools_controller.rb`: Handles CRUD operations for AI tools
c. Views and Components:
- New Ember.js components for tool management in the admin interface
- Updates to existing AI persona management components to support custom tools
d. Core functionality:
- `lib/ai_bot/tool_runner.rb`: Implements the custom tool execution system
- `lib/ai_bot/tools/custom.rb`: Defines the custom tool class
e. Routes and configurations:
- Updates to route configurations to include new AI tool management pages
f. Migrations:
- `db/migrate/20240618080148_create_ai_tools.rb`: Creates the ai_tools table
g. Tests:
- New test files for AI tool functionality and integration
The PR integrates the custom tools system with the existing AI persona framework, allowing personas to use both built-in and custom tools. It also includes safety measures such as timeouts and HTTP request limits to prevent misuse of custom tools.
Overall, this PR significantly enhances the flexibility and extensibility of the Discourse AI system by allowing administrators to create and manage custom AI tools tailored to their specific needs.
Co-authored-by: Martin Brennan <[email protected]>
Having this as a callback prevents deploys of sites with a vLLM SRV configured and pending migrations. Additionally, this fixes a bug where we didn't delete/deactivate the companion user after deleting an LLM.
Previously, we stored request parameters like the OpenAI organization and Bedrock's access key and region as site settings. This change stores them in the `llm_models` table instead, letting us drop more settings while also becoming more flexible.
* FEATURE: LLM presets for model creation
Previous to this users needed to look up complicated settings
when setting up models.
This introduces and extensible preset system with Google/OpenAI/Anthropic
presets.
This will cover all the most common LLMs, we can always add more as
we go.
Additionally:
- Proper support for Anthropic Claude Sonnet 3.5
- Stop blurring api keys when navigating away - this made it very complex to reuse keys
We no longer support the "provider:model" format in the "ai_helper_model" and
"ai_embeddings_semantic_search_hyde_model" settings. We'll migrate existing
values and work with our new data-driven LLM configs from now on.
Previously read tool only had access to public topics, this allows
access to all topics user has access to, if admin opts for the option
Also
- Fixes VLLM migration
- Display which llms have bot enabled
* DRAFT: Create AI Bot users dynamically and support custom LlmModels
* Get user associated to llm_model
* Track enabled bots with attribute
* Don't store bot username. Minor touches to migrate default values in settings
* Handle scenario where vLLM uses a SRV record
* Made 3.5-turbo-16k the default version so we can remove hack
- Display filtered search correctly, so it is not confusing
- When XML stripping, if a chunk was `<` it would crash
- SQL Helper improved to be better aware of Data Explorer
In some case we may be deploying migrations, seeding and then
running post migrations, we need this to work so we give up
on this small window of protection
This is a rather huge refactor with 1 new feature (tool details can
be suppressed)
Previously we use the name "Command" to describe "Tools", this unifies
all the internal language and simplifies the code.
We also amended the persona UI to use less DToggles which aligns
with our design guidelines.
Co-authored-by: Martin Brennan <[email protected]>
Native tools do not work well on Opus.
Chain of Thought prompting means it consumes enormous amounts of
tokens and has poor latency.
This commit introduce and XML stripper to remove various chain of
thought XML islands from anthropic prompts when tools are involved.
This mean Opus native tools is now functions (albeit slowly)
From local testing XML just works better now.
Also fixes enum support in Anthropic native tools
Add native Cohere tool support
- Introduce CohereTools class for tool translation and result processing
- Update Command dialect to integrate with CohereTools
- Modify Cohere endpoint to support passing tools and processing tool calls
- Add spec for testing tool triggering with Cohere endpoint
1. New tool to easily find files (and default branch) in a Github repo
2. Improved read tool with clearer params and larger context
* limit can totally mess up the richness semantic search adds, so include the results unconditionally.
Initial implementation allowed internet wide sharing of
AI conversations, on sites that require login.
This feature can be an anti feature for private sites cause they
can not share conversations internally.
For now we are removing support for public sharing on login required
sites, if the community need the feature we can consider adding a
setting.
Previoulsy on GPT-4-vision was supported, change introduces support
for Google/Anthropic and new OpenAI models
Additionally this makes vision work properly in dev environments
cause we sent the encoded payload via prompt vs sending urls
This change allows us to delete custom models. It checks if there is no module using them.
It also fixes a bug where the after-create transition wasn't working. While this prevents a model from being saved multiple times, endpoint validations are still needed (will be added in a separate PR).:
When triggering a PM from new-message route, we still had the UI
for "this is an official warning"
This removes that UI from bot messages, which is all clutter.
This is similar to code interpreter by ChatGPT, except that it uses
JavaScript as the execution engine.
Safeguards were added to ensure memory is constrained and evaluation
times out.
* FEATURE: Set endpoint credentials directly from LlmModel.
Drop Llama2Tokenizer since we no longer use it.
* Allow http for custom LLMs
---------
Co-authored-by: Rafael Silva <[email protected]>
- a post can be triaged a maximum of twice a minute
- system can run a total of 60 triages a minute
Low defaults were picked to safeguard against any possible loops
This can be amended if required via hidden site settings.
When lazy load categories is enabled, the list of categories does not
have to fetched from the "site.json" endpoint because it is already
returned by "search.json".
This commit reverts commits 5056502 and 3e54697 because iterating over
all pages of categories is not really necessary.
LLM selector control had no memory and was awkward to click.
Instead we now:
- Clearly display which llm you are talking to
- Allow you to change llm direct from composer
- Introduce new support for GPT4o (automation / bot / summary / helper)
- Properly account for token counts on OpenAI models
- Track feature that was used when generating AI completions
- Remove custom llm support for summarization as we need better interfaces to control registration and de-registration
There are still some limitations to which models we can support with the `LlmModel` class. This will enable support for Llama3 while we sort those out.
This PR introduces the concept of "LlmModel" as a new way to quickly add new LLM models without making any code changes. We are releasing this first version and will add incremental improvements, so expect changes.
The AI Bot can't fully take advantage of this feature as users are hard-coded. We'll fix this in a separate PR.s
The menu service doesn’t implement an activeMenu property anymore as it can now support concurrent menus. The solution to this is to use `getByIdentifier`.
This optional feature allows search to be performed in the context
of the user that executed it.
By default we do not allow this behavior cause it means llm gets
access to potentially secure data.
When the bot is @mentioned, we need to be a lot more careful
about constructing context otherwise bot gets ultra confused.
This changes multiple things:
1. We were omitting all thread first messages (fixed)
2. Include thread title (if available) in context
3. Construct context in a clearer way separating user request from data
Both endpoints provide OpenAI-compatible servers. The only difference is that Vllm doesn't support passing tools as a separate parameter. Even if the tool param is supported, it ultimately relies on the model's ability to handle native functions, which is not the case with the models we have today.
As a part of this change, we are dropping support for StableBeluga/Llama2 models. They don't have a chat_template, meaning the new API can translate them.
These changes let us remove some of our existing dialects and are a first step in our plan to support any LLM by defining them as data-driven concepts.
I rewrote the "translate" method to use a template method and extracted the tool support strategies into its classes to simplify the code.
Finally, these changes bring support for Ollama when running in dev mode. It only works with Mistral for now, but it will change soon..
* Well, it was quite a journey but now tools have "context" which
can be critical for the stuff they generate
This entire change was so Dall E and Artist generate images in the correct context
* FIX: improve error handling around image generation
- also corrects image markdown and clarifies code
* fix spec
Add support for chat with AI personas
- Allow enabling chat for AI personas that have an associated user
- Add new setting `allow_chat` to AI persona to enable/disable chat
- When a message is created in a DM channel with an allowed AI persona user, schedule a reply job
- AI replies to chat messages using the persona's `max_context_posts` setting to determine context
- Store tool calls and custom prompts used to generate a chat reply on the `ChatMessageCustomPrompt` table
- Add tests for AI chat replies with tools and context
At the moment unlike posts we do not carry tool calls in the context.
No @mention support yet for ai personas in channels, this is future work
The initial setup done in fb0d56324f
clashed with other plugins, I found this when trying to do the same
for Gamification. This uses a better routing setup and removes the
need to define the config nav link for Settings -- that is always inserted.
Relies on https://github.com/discourse/discourse/pull/26707
A recent change meant that llm instance got cached internally, repeat calls
to inference would cache data in Endpoint object leading model to
failures.
Both Gemini and Open AI expect a clean endpoint object cause they
set data.
This amends internals to make sure llm.generate will always operate
on clean objects
This commit introduces a new feature for AI Personas called the "Question Consolidator LLM". The purpose of the Question Consolidator is to consolidate a user's latest question into a self-contained, context-rich question before querying the vector database for relevant fragments. This helps improve the quality and relevance of the retrieved fragments.
Previous to this change we used the last 10 interactions, this is not ideal cause the RAG would "lock on" to an answer.
EG:
- User: how many cars are there in europe
- Model: detailed answer about cars in europe including the term car and vehicle many times
- User: Nice, what about trains are there in the US
In the above example "trains" and "US" becomes very low signal given there are pages and pages talking about cars and europe. This mean retrieval is sub optimal.
Instead, we pass the history to the "question consolidator", it would simply consolidate the question to "How many trains are there in the United States", which would make it fare easier for the vector db to find relevant content.
The llm used for question consolidator can often be less powerful than the model you are talking to, we recommend using lighter weight and fast models cause the task is very simple. This is configurable from the persona ui.
This PR also removes support for {uploads} placeholder, this is too complicated to get right and we want freedom to shift RAG implementation.
Key changes:
1. Added a new `question_consolidator_llm` column to the `ai_personas` table to store the LLM model used for question consolidation.
2. Implemented the `QuestionConsolidator` module which handles the logic for consolidating the user's latest question. It extracts the relevant user and model messages from the conversation history, truncates them if needed to fit within the token limit, and generates a consolidated question prompt.
3. Updated the `Persona` class to use the Question Consolidator LLM (if configured) when crafting the RAG fragments prompt. It passes the conversation context to the consolidator to generate a self-contained question.
4. Added UI elements in the AI Persona editor to allow selecting the Question Consolidator LLM. Also made some UI tweaks to conditionally show/hide certain options based on persona configuration.
5. Wrote unit tests for the QuestionConsolidator module and updated existing persona tests to cover the new functionality.
This feature enables AI Personas to better understand the context and intent behind a user's question by consolidating the conversation history into a single, focused question. This can lead to more relevant and accurate responses from the AI assistant.
Updating the editing model's rag_uploads in the editor component broke multi-file uploading. Instead, we'll keep the uploads in the uploader and update the model when we finish.
This PR also fast-tracks the initial update so we can show feedback to the user quickly, and allows uploading MD files.
Bug reported on https://meta.discourse.org/t/discourse-ai-persona-upload-support/304049/11
This allows you to exclude trees of categories in a simple way
It also means you can no longer exclude "just the parent" but
this is a fair compromise.
- Adds support for sd3 and sd3 turbo models - this requires new endpoints
- Adds a hack to normalize arrays in the tool calls
- Removes some leftover code
- Adds support for aspect ratio as well so you can generate wide or tall images
For quite a few weeks now, some times, when running function calls
on Anthropic we would get a "stray" - "calls" line.
This has been enormously frustrating!
I have been unable to find the source of the bug so instead decoupled
the implementation and create a very clear "function call normalizer"
This new class is extensively tested and guards against the type of
edge cases we saw pre-normalizer.
This also simplifies the implementation of "endpoint" which no longer
needs to handle all this complex logic.
- Updated AI Bot to only support Gemini 1.5 (used to support 1.0) - 1.0 was removed cause it is not appropriate for Bot usage
- Summaries and automation can now lean on Gemini 1.5 pro
- Amazon added support for Claude 3 Opus, added internal support for it on bedrock
* FIX: various RAG edge cases
- Nicer text to describe RAG, avoids the word RAG
- Do not attempt to save persona when removing uploads and it is not created
- Remove old code that avoided touching rag params on create
* FIX: Missing pause button for persona users
* Feature: allow specific users to debug ai request / response chains
This can help users easily tune RAG and figure out what is going
on with requests.
* discourse helper so it does not explode
* fix test
* simplify implementation
* FEATURE: allow tuning of RAG generation
- change chunking to be token based vs char based (which is more accurate)
- allow control over overlap / tokens per chunk and conversation snippets inserted
- UI to control new settings
* improve ui a bit
* fix various reindex issues
* reduce concurrency
* try ultra low queue ... concurrency 1 is too slow.
Just having the word JSON can confuse models when we expect them
to deal solely in XML
Instead provide an example of how string arrays should be returned
Technically the tool framework supports int arrays and more, but
our current implementation only does string arrays.
Also tune the prompt construction not to give any tips about arrays
if none exist
This commit uses a new plugin modifier introduced in https://github.com/discourse/discourse/pull/26508
to mark all uploads as _not_ secure in shared PM AI conversations.
This is so images created by the AI bot (or uploaded by the user)
do not end up as broken URLs because of the security requirements
around them.
This relies on the UpdateTopicUploadSecurity job in core as well,
which is fired when an AI conversation is shared or deleted.
- Added Cohere Command models (Command, Command Light, Command R, Command R Plus) to the available model list
- Added a new site setting `ai_cohere_api_key` for configuring the Cohere API key
- Implemented a new `DiscourseAi::Completions::Endpoints::Cohere` class to handle interactions with the Cohere API, including:
- Translating request parameters to the Cohere API format
- Parsing Cohere API responses
- Supporting streaming and non-streaming completions
- Supporting "tools" which allow the model to call back to discourse to lookup additional information
- Implemented a new `DiscourseAi::Completions::Dialects::Command` class to translate between the generic Discourse AI prompt format and the Cohere Command format
- Added specs covering the new Cohere endpoint and dialect classes
- Updated `DiscourseAi::AiBot::Bot.guess_model` to map the new Cohere model to the appropriate bot user
In summary, this PR adds support for using the Cohere Command family of models with the Discourse AI plugin. It handles configuring API keys, making requests to the Cohere API, and translating between Discourse's generic prompt format and Cohere's specific format. Thorough test coverage was added for the new functionality.
BAAI/bge-m3 is an interesting model, that is multilingual and with a
context size of 8192. Even with a 16x larger context, it's only 4x slower
to compute it's embeddings on the worst case scenario.
Also includes a minor refactor of the rake task, including setting model
and concurrency levels when running the backfill task.
Open AI just released gpt-4-turbo (with vision)
This change stops using the old preview model and swaps with the
officially released gpt-4-turbo
To come is an implementation of vision.
It used to fetch it from /site.json, but /categories.json is the more
appropriate one. This one also implements pagination, so we have to do
one request per page.
* FEATURE: Add metadata support for RAG
You may include non indexed metadata in the RAG document by using
[[metadata ....]]
This information is attached to all the text below and provided to
the retriever.
This allows for RAG to operate within a rich amount of contexts
without getting lost
Also:
- re-implemented chunking algorithm so it streams
- moved indexing to background low priority queue
* Baran gem no longer required.
* tokenizers is on 4.4 ... upgrade it ...
it is close in performance to GPT 4 at a fraction of the cost,
nice to add it to the mix.
Also improves a test case to simulate streaming, I am hunting for
the "calls" word that is jumping into function calls and can't quite
find it.
This PR lets you associate uploads to an AI persona, which we'll split and generate embeddings from. When building the system prompt to get a bot reply, we'll do a similarity search followed by a re-ranking (if available). This will let us find the most relevant fragments from the body of knowledge you associated with the persona, resulting in better, more informed responses.
For now, we'll only allow plain-text files, but this will change in the future.
Commits:
* FEATURE: RAG embeddings for the AI Bot
This first commit introduces a UI where admins can upload text files, which we'll store, split into fragments,
and generate embeddings of. In a next commit, we'll use those to give the bot additional information during
conversations.
* Basic asymmetric similarity search to provide guidance in system prompt
* Fix tests and lint
* Apply reranker to fragments
* Uploads filter, css adjustments and file validations
* Add placeholder for rag fragments
* Update annotations
This pull request makes several improvements and additions to the GitHub-related tools and personas in the `discourse-ai` repository:
1. It adds the `WebBrowser` tool to the `Researcher` persona, allowing the AI to visit web pages, retrieve HTML content, extract the main content, and convert it to plain text.
2. It updates the `GithubFileContent`, `GithubPullRequestDiff`, and `GithubSearchCode` tools to handle HTTP responses more robustly (introducing size limits).
3. It refactors the `send_http_request` method in the `Tool` class to follow redirects when specified, and to read the response body in chunks to avoid memory issues with large responses. (only for WebBrowser)
4. It updates the system prompt for the `Researcher` persona to provide more detailed guidance on when to use Google search vs web browsing, and how to optimize tool usage and reduce redundant requests.
5. It adds a new `web_browser_spec.rb` file with tests for the `WebBrowser` tool, covering various scenarios like handling different HTML structures and following redirects.
This commit adds the ability to enable vision for AI personas, allowing them to understand images that are posted in the conversation.
For personas with vision enabled, any images the user has posted will be resized to be within the configured max_pixels limit, base64 encoded and included in the prompt sent to the AI provider.
The persona editor allows enabling/disabling vision and has a dropdown to select the max supported image size (low, medium, high). Vision is disabled by default.
This initial vision support has been tested and implemented with Anthropic's claude-3 models which accept images in a special format as part of the prompt.
Other integrations will need to be updated to support images.
Several specs were added to test the new functionality at the persona, prompt building and API layers.
- Gemini is omitted, pending API support for Gemini 1.5. Current Gemini bot is not performing well, adding images is unlikely to make it perform any better.
- Open AI is omitted, vision support on GPT-4 it limited in that the API has no tool support when images are enabled so we would need to full back to a different prompting technique, something that would add lots of complexity
---------
Co-authored-by: Martin Brennan <[email protected]>
report runner and llm triage used different paths to figure out
underlying model name, unify so we use the same path.
fixes claude 3 based models on llm triage
Prompt was steering incorrectly into the wrong language.
New prompt attempts to be more concise and clear and provides
better guidance about size of summary and how to format it.
We were only suppressing non mentions, ones that become spans.
@sam in the test was not resolving to a mention cause the user
did not exist.
depends on: https://github.com/discourse/discourse/pull/26253 for tests to pass.
- Stop replying as bot, when human replies to another human
- Reply as correct persona when replying directly to a persona
- Fix paper cut where suppressing notifications was not doing so
This PR consolidates the implements new Anthropic Messages interface for Bedrock Claude endpoints and adds support for the new Claude 3 models (haiku, opus, sonnet).
Key changes:
- Renamed `AnthropicMessages` and `Anthropic` endpoint classes into a single `Anthropic` class (ditto for ClaudeMessages -> Claude)
- Updated `AwsBedrock` endpoints to use the new `/messages` API format for all Claude models
- Added `claude-3-haiku`, `claude-3-opus` and `claude-3-sonnet` model support in both Anthropic and AWS Bedrock endpoints
- Updated specs for the new consolidated endpoints and Claude 3 model support
This refactor removes support for old non messages API which has been deprecated by anthropic
* FEATURE: allow suppression of notifications from report generation
Previously we needed to do this by hand, unfortunately this uses up
too many tokens and is very hard to discover.
New option means that we can trivially disable notifications without
needing any prompt engineering.
* URI.parse is safer, use it
- Adds a nonce to both script tags
- Removes the `onload=` inline script, and moves the tags to the end of the `<body>` instead. This provides the same UX (page will load and render, then hljs will be applied when ready)
* FIX: Handle unicode on tokenizer
Our fast track code broke when strings had characters who are longer in tokens than
in UTF-8.
Admins can set `DISCOURSE_AI_STRICT_TOKEN_COUNTING: true` in app.yml to ensure token counting is strict, even if slower.
Co-authored-by: wozulong <[email protected]>
* FIX: don't show share conversation incorrectly
- ai_persona_name can be null vs undefined leading to button showing up where it should not
- do not allow sharing of conversations where user is sending PMs to self
* remove erroneous code
* avoid query
This allows users to share a static page of an AI conversation with
the rest of the world.
By default this feature is disabled, it is enabled by turning on
ai_bot_allow_public_sharing via site settings
Precautions are taken when sharing
1. We make a carbonite copy
2. We minimize work generating page
3. We limit to 100 interactions
4. Many security checks - including disallowing if there is a mix
of users in the PM.
* Bonus commit, large PRs like this PR did not work with github tool
large objects would destroy context
Co-authored-by: Martin Brennan <[email protected]>
Adds support for "name" on functions which can be used for tool calls
For function calls we need to keep track of id/name and previously
we only supported either
Also attempts to improve sql helper
This PR adds AI semantic search to the search pop available on every page.
It depends on several new and optional settings, like per post embeddings and a reranker model, so this is an experimental endeavour.
---------
Co-authored-by: Rafael Silva <[email protected]>
1. Fix input fields in AI persona editor and make GitHub tool authentication optional
2. AI persona editor improvements and tool GitHub access token check
This pull request makes a few improvements:
- Adds `lang="en"` to number input fields in the AI persona editor to prevent localization issues
- Adds `step="any"` to allow fractional values for temperature and top_p settings
- Makes GitHub tool authentication contingent on `ai_bot_github_access_token` site setting being present
see: https://meta.discourse.org/t/ai-bot-personas-don-t-accept-decimals-for-temperature-top-p/298243/7
Introduces a new AI Bot persona called 'GitHub Helper' which is specialized in assisting with GitHub-related tasks and questions. It includes the following key changes:
- Implements the GitHub Helper persona class with its system prompt and available tools
- Adds three new AI Bot tools for GitHub interactions:
- github_file_content: Retrieves content of files from a GitHub repository
- github_pull_request_diff: Retrieves the diff for a GitHub pull request
- github_search_code: Searches for code in a GitHub repository
- Updates the AI Bot dialects to support the new GitHub tools
- Implements multiple function calls for standard tool dialect
Chat thread replies draft trigger the thread_created event, which we relied on
to trigger the AI generated title. Because of that we now will use the noisier
chat_message_created event, and manually check for thread and replies existence.
See https://github.com/discourse/discourse/pull/26033
Previous pinned version included references to new plugin API not
available on stable.
This new pinned version also contains 'backported' SSRF protection
security fix.
This provides new support for messages API from Claude.
It is required for latest model access.
Also corrects implementation of function calls.
* Fix message interleving
* fix broken spec
* add new models to automation
- FIX: only update system attributes when updating system persona
- FIX: update participant count by hand so bot messages show in inbox
Co-authored-by: Joffrey JAFFEUX <[email protected]>
* FIX: support multiple tool calls
Prior to this change we had a hard limit of 1 tool call per llm
round trip. This meant you could not google multiple things at
once or perform searches across two tools.
Also:
- Hint when Google stops working
- Log topic_id / post_id when performing completions
* Also track id for title
Previous to this fix if a tool call ever streamed a SPACE alone,
we would eat it and ignore it, breaking params
Also fixes some tests to ensure they are actually called :)
* DEV: improve internal design of ai persona and bug fix
- Fixes bug where OpenAI could not describe images
- Fixes bug where mentionable personas could not be mentioned unless overarching bot was enabled
- Improves internal design of playground and bot to allow better for non "bot" users
- Allow PMs directly to persona users (previously bot user would also have to be in PM)
- Simplify internal code
Co-authored-by: Martin Brennan <[email protected]>
* FEATURE: AI helper support in non English languages
This attempts some prompt engineering to coerce AI helper to answer
in the appropriate language.
Note mileage will vary, in testing GPT-4 produces the best results
GPT-3.5 can return OKish results.
* Extend non english support for GPT-4V image caption
* Update db/fixtures/ai_helper/603_completion_prompts.rb
---------
Co-authored-by: Rafael Silva <[email protected]>
The Faraday adapter and `FinalDestionation::HTTP` will protect us from admin-initiated SSRF attacks when interacting with the external services powering this plugin features.:
Persona users are still bots, but we were not properly accounting
for it and share icon was not showing up.
This depends on a core change that adds .topic to transformed posts
Utilizes the check for secure upload permissions from core PR
https://github.com/discourse/discourse/pull/25758 and cleans up
controller codes and spec code to reuse existing code and better
reflect reality.
This PR adds a new feature where you can generate captions for images in the composer using AI.
---------
Co-authored-by: Rafael Silva <[email protected]>
This persona searches Discourse Meta for help with Discourse and
points users at relevant posts.
It is somewhat similar to using "Forum Helper" on meta, with the
notable difference that we can not lean on semantic search so using
some prompt engineering we try to keep it simple.
Affects the following settings:
ai_toxicity_groups_bypass
ai_helper_allowed_groups
ai_helper_custom_prompts_allowed_groups
post_ai_helper_allowed_groups
This turns off client: true for these group-based settings,
because there is no guarantee that the current user gets all
their group memberships serialized to the client. Better to check
server-side first.
1. Personas are now optionally mentionable, meaning that you can mention them either from public topics or PMs
- Mentioning from PMs helps "switch" persona mid conversation, meaning if you want to look up sites setting you can invoke the site setting bot, or if you want to generate an image you can invoke dall e
- Mentioning outside of PMs allows you to inject a bot reply in a topic trivially
- We also add the support for max_context_posts this allow you to limit the amount of context you feed in, which can help control costs
2. Add support for a "random picker" tool that can be used to pick random numbers
3. Clean up routing ai_personas -> ai-personas
4. Add Max Context Posts so users can control how much history a persona can consume (this is important for mentionable personas)
Co-authored-by: Martin Brennan <[email protected]>
* FIX: Better AI chat thread titles
- Fix quote removal when multi-line
- Use XML tags for better LLM output parsing
- Use stop_sequences for faster and less wasteful LLM calls
- Adds truncation as the last line of defense
* FEATURE: allow personas to supply top_p and temperature params
Code assistance generally are more focused at a lower temperature
This amends it so SQL Helper runs at 0.2 temperature vs the more
common default across LLMs of 1.0.
Reduced temperature leads to more focused, concise and predictable
answers for the SQL Helper
* fix tests
* This is not perfect, but far better than what we do today
Instead of fishing for
1. Draft sequence
2. Draft body
We skip (2), this means the composer "only" needs 1 http request to
open, we also want to eliminate (1) but it is a bit of a trickier
core change, may figure out how to pull it off (defer it to first draft save)
Value of bot drafts < value of opening bot conversations really fast
When bedrock rate limits it returns a 200 BUT also returns a JSON
document with the error.
Previously we had no special case here so we complained about nil
New code properly logs the problem
The idea is to increase the frequency so we can run with smaller batch sizes.
Big batches cause problems when running backups, so it's better to have shorter but
more frequent jobs.
1. on failure we were queuing a job to generate embeddings, it had the wrong params. This is both fixed and covered in a test.
2. backfill embedding in the order of bumped_at, so newest content is embedded first, cover with a test
3. add a safeguard for hidden site setting that only allows batches of 50k in an embedding job run
Previously old embeddings were updated in a random order, this changes it so we update in a consistent order
- Allow users to supply top_p and temperature values, which means people can fine tune randomness
- Fix bad localization string
- Fix bad remapping of max tokens in gemini
- Add support for top_p as a general param to llms
- Amend system prompt so persona stops treating a user as an adversary
* UX: Validations to Llm-backed features (except AI Bot)
This change is part of an ongoing effort to prevent enabling a broken feature due to lack of configuration. We also want to explicit which provider we are going to use. For example, Claude models are available through AWS Bedrock and Anthropic, but the configuration differs.
Validations are:
* You must choose a model before enabling the feature.
* You must turn off the feature before setting the model to blank.
* You must configure each model settings before being able to select it.
* Add provider name to summarization options
* vLLM can technically support same models as HF
* Check we can talk to the selected model
* Check for Bedrock instead of anthropic as a site could have both creds setup
* FEATURE: add support for new OpenAI embedding models
This adds support for just released text_embedding_3_small and large
Note, we have not yet implemented truncation support which is a
new API feature. (triggered using dimensions)
* Tiny side fix, recalc bots when ai is enabled or disabled
* FIX: downsample to 2000 items per vector which is a pgvector limitation
We were not validating input for generate leading to 2 tests not
failing correctly despite functionality being broken.
This ensures that input is validated,and in turn fixes the broken
specs
When you trim a prompt we never want to have a state where there
is a "tool" reply without a corresponding tool call, it makes no
sense
Also
- GPT-4-Turbo is 128k, fix that
- Claude was not preserving username in prompt
- We were throwing away unicode usernames instead of adding to
message
We're updating core to change TL based access settings to be group based. This requires some updates of tests to work correctly. (The existing test setup gives false positives.)
Account properly for function calls, don't stream through <details> blocks
- Rush cooked content back to client
- Wait longer (up to 60 seconds) before giving up on streaming
- Clean up message bus channels so we don't have leftover data
- Make ai streamer much more reusable and much easier to read
- If buffer grows quickly, rush update so you are not artificially waiting
- Refine prompt interface
- Fix lost system message when prompt gets long
* REFACTOR: Represent generic prompts with an Object.
* Adds a bit more validation for clarity
* Rewrite bot title prompt and fix quirk handling
---------
Co-authored-by: Sam Saffron <[email protected]>
Fixes a regression from 140359c which caused we to set this globally based on post count, rendering the cost of an index scan on the topics table too high and making the planner, correctly, not use the index anymore.
Hopefully https://github.com/pgvector/pgvector/issues/235 lands soon.
This PR introduces 3 things:
1. Fake bot that can be used on local so you can test LLMs, to enable on dev use:
SiteSetting.ai_bot_enabled_chat_bots = "fake"
2. More elegant smooth streaming of progress on LLM completion
This leans on JavaScript to buffer and trickle llm results through. It also amends it so the progress dot is much
more consistently rendered
3. It fixes the Claude dialect
Claude needs newlines **exactly** at the right spot, amended so it is happy
---------
Co-authored-by: Martin Brennan <[email protected]>
Followup 2636efcd1b,
whenever ruby code was changed locally this would break
module loading, giving an "uninitialized constant
DiscourseAi::Embeddings::EntryPoint::SemanticRelated" error.
This allows admins to configure services with multiple backends using DNS SRV records. This PR also adds support for shared secret auth via headers for TEI and vLLM endpoints, so they are inline with the other ones.
Previous to this change it was very hard to tell if completion was
stuck or not.
This introduces a "dot" that follows the completion and starts
flashing after 5 seconds.
* FIX: improve bot behavior
- Provide more information to Gemini context post function execution
- Use system prompts for Claude (fixes Dall E)
- Ensure Assistant is properly separated
- Teach Claude to return arrays in JSON vs XML
Also refactors tests so we do not copy tool preamble everywhere
* System msg is claude-2 only. fix typo
---------
Co-authored-by: Roman Rizzi <[email protected]>
We thought Azure's latest API version didn't have tool support yet, but I didn't understand it was complaining about a required field in the tool call message.
* FIX: don't include <details> in context
We need to be careful adding <details> into context of conversations
it can cause LLMs to hallucinate results
* Fix Gemini multi-turn ctx flattening
---------
Co-authored-by: Roman Rizzi <[email protected]>
DALL E command accepts an Array as a tool argument, this was not
parsed correctly by the invoker leading to errors generating
images with DALL E
Side quest ... don't use update! it calls validations and will now
fail due to email validation
It also corrects the syntax around tool support, which was wrong.
Gemini doesn't want us to include messages about previous tool invocations, so I had to shuffle around some code to send the response it generated from those invocations instead. For this, I created the "multi_turn" context, which bundles all the context involved in the interaction.
* DEV: AI bot migration to the Llm pattern.
We added tool and conversation context support to the Llm service in discourse-ai#366, meaning we met all the conditions to migrate this module.
This PR migrates to the new pattern, meaning adding a new bot now requires minimal effort as long as the service supports it. On top of this, we introduce the concept of a "Playground" to separate the PM-specific bits from the completion, allowing us to use the bot in other contexts like chat in the future. Commands are called tools, and we simplified all the placeholder logic to perform updates in a single place, making the flow more one-wayish.
* Followup fixes based on testing
* Cleanup unused inference code
* FIX: text-based tools could be in the middle of a sentence
* GPT-4-turbo support
* Use new LLM API
* FIX: AI helper not working correctly with mixtral
This PR introduces a new function on the generic llm called #generate
This will replace the implementation of completion!
#generate introduces a new way to pass temperature, max_tokens and stop_sequences
Then LLM implementers need to implement #normalize_model_params to
ensure the generic names match the LLM specific endpoint
This also adds temperature and stop_sequences to completion_prompts
this allows for much more robust completion prompts
* port everything over to #generate
* Fix translation
- On anthropic this no longer throws random "This is your translation:"
- On mixtral this actually works
* fix markdown table generation as well
Chrome and Firefox work with standard clipboardCopy, but since
we are making an ajax call Safari fails cause there is too much
delay detected.
To avoid this issue we trade in promises which are acceptable and
work in iOS.
Currently we're seeing 500s when related_topics are getting rendered. We should get the topic's category rather than on the array.
```
ActionView::Template::Error (undefined method `category' for [#<Topic id ... ]
```
* FEATURE: allow easy sharing of bot conversations
* Lean on new core API i
* Added system spec for copy functionality
* Update assets/javascripts/initializers/ai-bot-replies.js
Co-authored-by: Alan Guo Xiang Tan <[email protected]>
* discourse later insted of setTimeout
* Update spec/system/ai_bot/share_spec.rb
Co-authored-by: Alan Guo Xiang Tan <[email protected]>
* feedback from review
just check the whole payload
* remove uneeded code
* fix spec
---------
Co-authored-by: Alan Guo Xiang Tan <[email protected]>
Previously endpoint/base would `+=` decoded_chunk to leftover
This could lead to cases where the leftover buffer had duplicate
previously processed data
Fix ensures we properly skip previously decoded data.
Introduce a Discourse Automation based periodical report. Depends on Discourse Automation.
Report works best with very large context language models such as GPT-4-Turbo and Claude 2.
- Introduces final_insts to generic llm format, for claude to work best it is better to guide the last assistant message (we should add this to other spots as well)
- Adds GPT-4 turbo support to generic llm interface
This PR adds tool support to available LLMs. We'll buffer tool invocations and return them instead of making users of this service parse the response.
It also adds support for conversation context in the generic prompt. It includes bot messages, user messages, and tool invocations, which we'll trim to make sure it doesn't exceed the prompt limit, then translate them to the correct dialect.
Finally, It adds some buffering when reading chunks to handle cases when streaming is extremely slow.:M
In https://github.com/discourse/discourse/pull/24740, `min_trust_to_create_topic` site setting was replaced by `create_topic_allowed_groups`. This PR replaces the former, deprecated one, with the latter.
This is somewhat experimental, but the context of likes/view/username
can help the llm find out what content is more important or even
common users that produce great content
This inflates the amount of tokens somewhat, but given it is all numbers
and search columns titles are only included once this is not severe
We were limiting to 20 results unconditionally cause we had to make
sure search always fit in an 8k context window.
Models such as GPT 3.5 Turbo (16k) and GPT 4 Turbo / Claude 2.1 (over 150k)
allow us to return a lot more results.
This means we have a much richer understanding cause context is far
larger.
This also allows a persona to tweak this number, in some cases admin
may want to be conservative and save on tokens by limiting results
This also tweaks the `limit` param which GPT-4 liked to set to tell
model only to use it when it needs to (and describes default behavior)
Keep in mind:
- GPT-4 is only going to be fully released next year - so this hardcodes preview model for now
- Fixes streaming bugs which became a big problem with GPT-4 turbo
- Adds Azure endpoing for turbo as well
Co-authored-by: Martin Brennan <[email protected]>
Personas now support providing options for commands.
This PR introduces a single option "base_query" for the SearchCommand. When supplied all searches the persona will perform will also include the pre-supplied filter.
This can allow personas to search a subset of the forum (such as documentation)
This system is extensible we can add options to any command trivially.
c.f. de983796e1b66aa2ab039a4fb6e32cec8a65a098
There will soon be additional login_required checks
for Guardian, and the intent of many checks by automated
systems is better fulfilled by using BasicUser, which
simulates a logged in TL0 forum user, rather than an
anon user.
* FEATURE: User sentiment on profile summary page
This introduces a new user stat in a user profile summary page.
It will show either neutral/positive/negative according to the dominant
sentiment in the user last interactions.
The user-stat widget is only rendered for staff.
Co-authored-by: Keegan George <[email protected]>
Previous to this change we relied on explicit loading for a files in Discourse AI.
This had a few downsides:
- Busywork whenever you add a file (an extra require relative)
- We were not keeping to conventions internally ... some places were OpenAI others are OpenAi
- Autoloader did not work which lead to lots of full application broken reloads when developing.
This moves all of DiscourseAI into a Zeitwerk compatible structure.
It also leaves some minimal amount of manual loading (automation - which is loading into an existing namespace that may or may not be there)
To avoid needing /lib/discourse_ai/... we mount a namespace thus we are able to keep /lib pointed at ::DiscourseAi
Various files were renamed to get around zeitwerk rules and minimize usage of custom inflections
Though we can get custom inflections to work it is not worth it, will require a Discourse core patch which means we create a hard dependency.
We must ensure we can isolate titles, and the models sometimes ignore the example we give them.
Additionally, anons can generate HyDE posts, so we need to check if user is nil when attempting to log requests.
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Joffrey JAFFEUX
Rafael dos Santos Silva
Keegan George
Roman Rizzi
David Taylor
Natalie Tay
Sam
Kelv
Ted Johansson
Discourse Translator Bot
Martin Brennan
Kris
Jarek Radosz
dependabot[bot]
David Battersby
Hoa Nguyen
Penar Musaraj
我秦始皇
Mark VanLandingham
Guhyoun Nam
Krzysztof Kotlarek
Tobias Eigen
Sérgio Saquetim
Osama Sayegh
Bianca Nenciu
jbrw
Loïc Guitaut
Hoa Nguyen
chapoi
Jan Cernik
PangBo
Régis Hanol
Selase Krakani
Dax74
Gerhard Schlager
Mal Curtis
Jordan Vidrine