Fix Database Schema

[0x4007]

This requires changes to the plugin that captures the data, as well as the database itself.

This must capture issue bodies.

---

Perhaps I wasn't explicit with my vision for the database schema, but this needs adjustments.

We only need:

id, created_at, modified_at, author_id, plaintext, embedding

• The plaintext should only save if its a public repo, otherwise we have liabilities on our hands storing company private data.
• We can make a separate table for when we start capturing messages from Telegram and other platforms. No concerns for now.
• ID might need to be the GraphQL node id, because, as I understand, the GitHub REST API will not allow us to query the comment with only the numerical ID. We would need to store the organization name, and the repository name, before we can query the comment based on its ID.
• Dates: the reason for capturing the dates is because I envision that we can make a Telegram chatbot that will allow anybody to query anything about the organization's latest projects in plain English. This is useful for onboarding (new team member has some questions about our infrastructure/products) and investors wanting an update. We need to give less weight to older comments.
• author id: for onboarding developers especially, it would be useful to target specific contributors' inputs. Mine for final decisions, and others who may be the primary contributors to specific repos, compared to other contributors etc.

[sshivaditya2019]

@0x4007 would plaintext include issue body as well or would it just be comment body?

So, right now if there's a private repository then the plaintext would become 'CENSORED'. Is that fine?

[0x4007]

Issue body as well and no leave null

[sshivaditya2019]

Issue body as well and no leave null

So, plaintext's value would issue body + comment body, or do we require a separate column, if that's the case then how should the embedding value be found? Is it just comment body or is it comment body + issue body.

[0x4007]

Why do you want to associate and/or combine the embedding of the issue body and the comment body?

Can you provide some examples of applications that would require this?

[sshivaditya2019]

Why do you want to associate and/or combine the embedding of the issue body and the comment body?

Can you provide some examples of applications that would require this?

Case 1: Issue Body + Comment Body:

• Only one vector index for Issue Body + Comment Body
• Might support vector search if we want to search for comments or issues alone.
• Uses: Relevance of the comment to the issue body, Text Based Searching using LIKE

Case 2: Issue Body and Comment Body Separate columns:

• Multiple Vector index required for vector search, as we want to support vector search by issue and comment text
• Multiple Columns (Might not be bad?)
• Uses: TLDR comments, comments/issue based vector search

If having two separate embeddings is acceptable, then that's fine. According to the issue's specification, the desired schema was id, created_at, modified_at, author_id, plaintext, embedding.

[0x4007]

I don't understand your explanation.

Anyways would be great if we can make the following applications:

1. Issue deduplication.
2. Assistant to answer questions specifically related to our DAO and products here on github and telegram in plain English

As I understand with my proposed schema, these applications are possible.

I wonder if it makes sense simply to capture every property, excluding the URL properties from the issue and comment objects on GitHub

[sshivaditya2019]

I don't understand your explanation.

Anyways would be great if we can make the following applications:

1. Issue deduplication.
2. Assistant to answer questions specifically related to our DAO and products here on github and telegram in plain English

As I understand with my proposed schema, these applications are possible.

I wonder if it makes sense simply to capture every property, excluding the URL properties from the issue and comment objects on GitHub

Just to be clear, what is plaintext in your schema ?

[0x4007]

The source code of the markdown of the comments.

I hope that the mark down syntax doesn't negatively affect the LLM's ability to understand.

I know that I always must clarify with ChatGPT that the right arrow bracket means it's a block quote from another comment.

I suppose we'll need to compare performance of raw marked down source code and something preprocessed/more similar to how we perceive it

[sshivaditya2019]

The source code of the markdown of the comments.

I hope that the mark down syntax doesn't negatively affect the LLM's ability to understand.

I know that I always must clarify with ChatGPT that the right arrow bracket means it's a block quote from another comment.

I suppose we'll need to compare performance of raw marked down source code and something preprocessed/more similar to how we perceive it

I'll give an example of plaintext let me know if this right ?

## Issue #1: Sample Issue with Replies

This is the body of the first issue.

### Comment by user1:

This is the first comment on issue #1.

#### Reply by user2:

This is a reply to the first comment by user1.

#### Reply by user3:

This is another reply to the first comment by user1.

### Comment by user4:

This is a second comment on issue #1.

Is this right ? If not could you please give an example. I am not sure what you mean by source code of the markdown of the comments.

[0x4007]

I am not an expert working with embeddings but as I understand we need to generate a single embedding per corpus of text (comment)

Otherwise yes you included the markdown syntax which is what I said.

[sshivaditya2019]

I am not an expert working with embeddings but as I understand we need to generate a single embedding per corpus of text (comment)

Otherwise yes you included the markdown syntax which is what I said.

To clarify, the example I shared involves a single embedding/vector. Currently, this means that a new text corpus(markdown I shared before) will be created for each new issue, and additional comments will be added to that corpus.

Currently, we create a new record for each comment and generate an embedding for each one. This approach won't work with the issuededup or Assistant plan you have.

[sshivaditya2019]

I am not an expert working with embeddings but as I understand we need to generate a single embedding per corpus of text (comment)

Otherwise yes you included the markdown syntax which is what I said.

So, to clarify, should the embedding be created per comment while the plaintext contains the entire markdown syntax I mentioned earlier? Is that the correct approach?

I don't think this will work as intended. We need either the comments with their embeddings or the text corpus (markdown) with its embeddings. Please let me know if I've misunderstood anything.

[0x4007]

So then if we are clobbering a single embedding to represent an entire issue (and pull) won't the costs increase exponentially? Can't we feed multiple embeddings to an LLM to work with?

It doesn't seem like a good approach to clobber

[sshivaditya2019]

So then if we are clobbering a single embedding to represent an entire issue (and pull) won't the costs increase exponentially? Can't we feed multiple embeddings to an LLM to work with?

It doesn't seem like a good approach to clobber

An alternative approach would be to store each piece of text from issues/comments as separate entities and associate them with their respective node IDs. This would create a large text corpus containing all the content from the organization, which could be identified by a global node ID or by using a type column in the schema.

[sshivaditya2019]

So then if we are clobbering a single embedding to represent an entire issue (and pull) won't the costs increase exponentially? Can't we feed multiple embeddings to an LLM to work with?

It doesn't seem like a good approach to clobber

To clarify, we don’t provide embeddings directly to the LLM. Instead, we perform a vector or similarity search, apply a ranking technique, and then extract text values that are most relevant to the search subject. This extracted text is then used as context for the LLM.

[0x4007]

@gentlementlegen im assuming storing node id won't be sufficient for us to determine which embeddings are associated with eachother. I'm assuming we need to also store the issue id

Anyways at this point im starting to think we should just store everything besides the _url properties and then we can optimize later since we seem to not know for sure what to do now.

[0x4007]

What are embeddings? OpenAI’s text embeddings measure the relatedness of text strings. Embeddings are commonly used for:

Search (where results are ranked by relevance to a query string) Clustering (where text strings are grouped by similarity) Recommendations (where items with related text strings are recommended) Anomaly detection (where outliers with little relatedness are identified) Diversity measurement (where similarity distributions are analyzed) Classification (where text strings are classified by their most similar label)

---

ChatGPT: Here are additional use cases for embeddings:

1. Sentiment Analysis: Embeddings can help map text to a sentiment space, allowing you to analyze the emotional tone of a text (positive, negative, neutral).
2. Language Translation: Embeddings capture semantic meaning, making them useful for machine translation, where similar sentences in different languages are close in embedding space.
3. Information Extraction: Identify and extract structured information from unstructured text by mapping words or phrases to relevant concepts.
4. Text Summarization: Summarize large documents by comparing sentence embeddings and extracting the most important, representative sentences.
5. Question Answering: Embeddings can improve the ability to match questions with the most relevant answers based on semantic similarity.
6. Text Generation: Used in generative models like GPT, embeddings help maintain coherence in text generation by representing context.
7. Content Moderation: Classify or filter harmful or inappropriate content by mapping texts to predefined categories of acceptable and non-acceptable behavior.
8. Topic Modeling: Discover underlying topics within a large set of documents by clustering similar text embeddings.

---

Reviewing the leaderboard and I realize that STS (Semantic Textual Similarity) is the most important for depuplication and task matchmaking. Check the STS tab. voyage-lite-02-instruct looks promising (the first ranked seems specifically designed for English-French translations?)

Seems significantly better than OpenAI's offering.

We may also be able to make a feature that allows contributors to ask product/strategic direction questions to the bot. It can link back to comments that answer their question.

Text summarization might be able to improve our relevance scoring of comments (verbose and concise comments may be able to receive similar "quantitative scoring" credit)

[sshivaditya2019]

@gentlementlegen im assuming storing node id won't be sufficient for us to determine which embeddings are associated with eachother. I'm assuming we need to also store the issue id

Anyways at this point im starting to think we should just store everything besides the _url properties and then we can optimize later since we seem to not know for sure what to do now.

@0x4007

Hi, could you update me on the final database schema? I believe that the global node ID and type should be sufficient. Other parameters might not be relevant to the embedding text corpus and could end up being more of a log dump.

[sshivaditya2019]

What are embeddings? OpenAI’s text embeddings measure the relatedness of text strings. Embeddings are commonly used for: Search (where results are ranked by relevance to a query string) Clustering (where text strings are grouped by similarity) Recommendations (where items with related text strings are recommended) Anomaly detection (where outliers with little relatedness are identified) Diversity measurement (where similarity distributions are analyzed) Classification (where text strings are classified by their most similar label)

ChatGPT: Here are additional use cases for embeddings:

1. Sentiment Analysis: Embeddings can help map text to a sentiment space, allowing you to analyze the emotional tone of a text (positive, negative, neutral).
2. Language Translation: Embeddings capture semantic meaning, making them useful for machine translation, where similar sentences in different languages are close in embedding space.
3. Information Extraction: Identify and extract structured information from unstructured text by mapping words or phrases to relevant concepts.
4. Text Summarization: Summarize large documents by comparing sentence embeddings and extracting the most important, representative sentences.
5. Question Answering: Embeddings can improve the ability to match questions with the most relevant answers based on semantic similarity.
6. Text Generation: Used in generative models like GPT, embeddings help maintain coherence in text generation by representing context.
7. Content Moderation: Classify or filter harmful or inappropriate content by mapping texts to predefined categories of acceptable and non-acceptable behavior.
8. Topic Modeling: Discover underlying topics within a large set of documents by clustering similar text embeddings.

Reviewing the leaderboard and I realize that STS (Semantic Textual Similarity) is the most important for depuplication and task matchmaking. Check the STS tab. voyage-lite-02-instruct looks promising (the first ranked seems specifically designed for English-French translations?)

Seems significantly better than OpenAI's offering.

We may also be able to make a feature that allows contributors to ask product/strategic direction questions to the bot. It can link back to comments that answer their question.

Text summarization might be able to improve our relevance scoring of comments (verbose and concise comments may be able to receive similar "quantitative scoring" credit)

I can probably write a new adapter for them, but I think their free tier has only 3 RPM (Request per minute), but offer 50M free tokens for new accounts to use ?

[0x4007]

Three requests per minute seems acceptable as long as we can have some type of buffer or queue. I believe we had plans to do this anyways with the OpenAI embeddings because it's significantly cheaper.

However, can we generate multiple embeddings per request? If not then I imagine this can get more complicated than a cron job.

As for schema, I think it's simpler to dump everything and optimize later. This is because we don't know what the next plugins will need and this gives us more flexibility for research. To me it's only a priority to optimize when we start hitting free tier limits.

[sshivaditya2019]

Three requests per minute seems acceptable as long as we can have some type of buffer or queue. I believe we had plans to do this anyways with the OpenAI embeddings because it's significantly cheaper.

However, can we generate multiple embeddings per request? If not then I imagine this can get more complicated than a cron job.

As for schema, I think it's simpler to dump everything and optimize later. This is because we don't know what the next plugins will need and this gives us more flexibility for research. To me it's only a priority to optimize when we start hitting free tier limits.

So, if we add a payment method, the requests per minute can be increased. I have updated the plugin for support with voyageai

[0x4007]

Any name and zip code should work

[sshivaditya2019]

@0x4007

I've added my payment method. Please delete this information immediately just to be safe or share it through another platform, like Telegram.

[0x4007]

It's merchant locked for up to $1

[sshivaditya2019]

So, the final schema would be Id (GlobalNodeId), plaintext(text from the comment),createdAt, ModifiedAt, embedding and this entire thing as a serialized object

type Comment = {
        /**
         * AuthorAssociation
         * @description How the author is associated with the repository.
         * @enum {string}
         */
        author_association:
          | "COLLABORATOR"
          | "CONTRIBUTOR"
          | "FIRST_TIMER"
          | "FIRST_TIME_CONTRIBUTOR"
          | "MANNEQUIN"
          | "MEMBER"
          | "NONE"
          | "OWNER";
        /** @description Contents of the issue comment */
        body: string;
        /** Format: date-time */
        created_at: string;
        /** Format: uri */
        html_url: string;
        /**
         * Format: int64
         * @description Unique identifier of the issue comment
         */
        id: number;
        /** Format: uri */
        issue_url: string;
        node_id: string;
        performed_via_github_app: null | components["schemas"]["integration"];
        /** Reactions */
        reactions: {
          "+1": number;
          "-1": number;
          confused: number;
          eyes: number;
          heart: number;
          hooray: number;
          laugh: number;
          rocket: number;
          total_count: number;
          /** Format: uri */
          url: string;
        };
        /** Format: date-time */
        updated_at: string;
        /**
         * Format: uri
         * @description URL for the issue comment
         */
        url: string;
        /** User */
        user: {
          /** Format: uri */
          avatar_url?: string;
          deleted?: boolean;
          email?: string | null;
          /** Format: uri-template */
          events_url?: string;
          /** Format: uri */
          followers_url?: string;
          /** Format: uri-template */
          following_url?: string;
          /** Format: uri-template */
          gists_url?: string;
          gravatar_id?: string;
          /** Format: uri */
          html_url?: string;
          id: number;
          login: string;
          name?: string;
          node_id?: string;
          /** Format: uri */
          organizations_url?: string;
          /** Format: uri */
          received_events_url?: string;
          /** Format: uri */
          repos_url?: string;
          site_admin?: boolean;
          /** Format: uri-template */
          starred_url?: string;
          /** Format: uri */
          subscriptions_url?: string;
          /** @enum {string} */
          type?: "Bot" | "User" | "Organization";
          /** Format: uri */
          url?: string;
        } | null;
      };

Is this schema fine, or is there anything else to be added or removed ?

[0x4007]

Looks like there isn't a lot to save from there if we remove the URLs, nested objects like reactions and user details.

I'm surprised that's all that's there.

Performed via GitHub app should also be removed because we shouldn't be generating embeddings for bot comments.

So after all of that, looks like basically we might only be adding author association.

[sshivaditya2019]

Looks like there isn't a lot to save from there if we remove the URLs, nested objects like reactions and user details.

I'm surprised that's all that's there.

Performed via GitHub app should also be removed because we shouldn't be generating embeddings for bot comments.

I took this from octokit's type definition. I think reactions should be kept, as it could be a metric of engagement for comments with abnormally low relevance.

Also, should the created_at in the schema be referenced from the comment or when the actual record was created.

This is the entire dump of the webhook event

/** issue_comment created event */
    "webhook-issue-comment-created": {
      /** @enum {string} */
      action: "created";
      /**
       * issue comment
       * @description The [comment](https://docs.github.com/rest/issues/comments#get-an-issue-comment) itself.
       */
      comment: {
        /**
         * AuthorAssociation
         * @description How the author is associated with the repository.
         * @enum {string}
         */
        author_association:
          | "COLLABORATOR"
          | "CONTRIBUTOR"
          | "FIRST_TIMER"
          | "FIRST_TIME_CONTRIBUTOR"
          | "MANNEQUIN"
          | "MEMBER"
          | "NONE"
          | "OWNER";
        /** @description Contents of the issue comment */
        body: string;
        /** Format: date-time */
        created_at: string;
        /** Format: uri */
        html_url: string;
        /**
         * Format: int64
         * @description Unique identifier of the issue comment
         */
        id: number;
        /** Format: uri */
        issue_url: string;
        node_id: string;
        performed_via_github_app: null | components["schemas"]["integration"];
        /** Reactions */
        reactions: {
          "+1": number;
          "-1": number;
          confused: number;
          eyes: number;
          heart: number;
          hooray: number;
          laugh: number;
          rocket: number;
          total_count: number;
          /** Format: uri */
          url: string;
        };
        /** Format: date-time */
        updated_at: string;
        /**
         * Format: uri
         * @description URL for the issue comment
         */
        url: string;
        /** User */
        user: {
          /** Format: uri */
          avatar_url?: string;
          deleted?: boolean;
          email?: string | null;
          /** Format: uri-template */
          events_url?: string;
          /** Format: uri */
          followers_url?: string;
          /** Format: uri-template */
          following_url?: string;
          /** Format: uri-template */
          gists_url?: string;
          gravatar_id?: string;
          /** Format: uri */
          html_url?: string;
          id: number;
          login: string;
          name?: string;
          node_id?: string;
          /** Format: uri */
          organizations_url?: string;
          /** Format: uri */
          received_events_url?: string;
          /** Format: uri */
          repos_url?: string;
          site_admin?: boolean;
          /** Format: uri-template */
          starred_url?: string;
          /** Format: uri */
          subscriptions_url?: string;
          /** @enum {string} */
          type?: "Bot" | "User" | "Organization";
          /** Format: uri */
          url?: string;
        } | null;
      };
      installation?: components["schemas"]["simple-installation"];
      /** @description The [issue](https://docs.github.com/rest/issues/issues#get-an-issue) the comment belongs to. */
      issue: {
        /** @enum {string|null} */
        active_lock_reason:
          | "resolved"
          | "off-topic"
          | "too heated"
          | "spam"
          | null;
        /** User */
        assignee?: {
          /** Format: uri */
          avatar_url?: string;
          deleted?: boolean;
          email?: string | null;
          /** Format: uri-template */
          events_url?: string;
          /** Format: uri */
          followers_url?: string;
          /** Format: uri-template */
          following_url?: string;
          /** Format: uri-template */
          gists_url?: string;
          gravatar_id?: string;
          /** Format: uri */
          html_url?: string;
          id: number;
          login: string;
          name?: string;
          node_id?: string;
          /** Format: uri */
          organizations_url?: string;
          /** Format: uri */
          received_events_url?: string;
          /** Format: uri */
          repos_url?: string;
          site_admin?: boolean;
          /** Format: uri-template */
          starred_url?: string;
          /** Format: uri */
          subscriptions_url?: string;
          /** @enum {string} */
          type?: "Bot" | "User" | "Organization" | "Mannequin";
          /** Format: uri */
          url?: string;
        } | null;
        assignees: ({
          /** Format: uri */
          avatar_url?: string;
          deleted?: boolean;
          email?: string | null;
          /** Format: uri-template */
          events_url?: string;
          /** Format: uri */
          followers_url?: string;
          /** Format: uri-template */
          following_url?: string;
          /** Format: uri-template */
          gists_url?: string;
          gravatar_id?: string;
          /** Format: uri */
          html_url?: string;
          id: number;
          login: string;
          name?: string;
          node_id?: string;
          /** Format: uri */
          organizations_url?: string;
          /** Format: uri */
          received_events_url?: string;
          /** Format: uri */
          repos_url?: string;
          site_admin?: boolean;
          /** Format: uri-template */
          starred_url?: string;
          /** Format: uri */
          subscriptions_url?: string;
          /** @enum {string} */
          type?: "Bot" | "User" | "Organization" | "Mannequin";
          /** Format: uri */
          url?: string;
        } | null)[];
        /**
         * AuthorAssociation
         * @description How the author is associated with the repository.
         * @enum {string}
         */
        author_association:
          | "COLLABORATOR"
          | "CONTRIBUTOR"
          | "FIRST_TIMER"
          | "FIRST_TIME_CONTRIBUTOR"
          | "MANNEQUIN"
          | "MEMBER"
          | "NONE"
          | "OWNER";
        /** @description Contents of the issue */
        body: string | null;
        /** Format: date-time */
        closed_at: string | null;
        comments: number;
        /** Format: uri */
        comments_url: string;
        /** Format: date-time */
        created_at: string;
        draft?: boolean;
        /** Format: uri */
        events_url: string;
        /** Format: uri */
        html_url: string;
        /** Format: int64 */
        id: number;
        labels?: {
          /** @description 6-character hex code, without the leading #, identifying the color */
          color: string;
          default: boolean;
          description: string | null;
          id: number;
          /** @description The name of the label. */
          name: string;
          node_id: string;
          /**
           * Format: uri
           * @description URL for the label
           */
          url: string;
        }[];
        /** Format: uri-template */
        labels_url: string;
        locked?: boolean;
        /**
         * Milestone
         * @description A collection of related issues and pull requests.
         */
        milestone: {
          /** Format: date-time */
          closed_at: string | null;
          closed_issues: number;
          /** Format: date-time */
          created_at: string;
          /** User */
          creator: {
            /** Format: uri */
            avatar_url?: string;
            deleted?: boolean;
            email?: string | null;
            /** Format: uri-template */
            events_url?: string;
            /** Format: uri */
            followers_url?: string;
            /** Format: uri-template */
            following_url?: string;
            /** Format: uri-template */
            gists_url?: string;
            gravatar_id?: string;
            /** Format: uri */
            html_url?: string;
            id: number;
            login: string;
            name?: string;
            node_id?: string;
            /** Format: uri */
            organizations_url?: string;
            /** Format: uri */
            received_events_url?: string;
            /** Format: uri */
            repos_url?: string;
            site_admin?: boolean;
            /** Format: uri-template */
            starred_url?: string;
            /** Format: uri */
            subscriptions_url?: string;
            /** @enum {string} */
            type?: "Bot" | "User" | "Organization" | "Mannequin";
            /** Format: uri */
            url?: string;
          } | null;
          description: string | null;
          /** Format: date-time */
          due_on: string | null;
          /** Format: uri */
          html_url: string;
          id: number;
          /** Format: uri */
          labels_url: string;
          node_id: string;
          /** @description The number of the milestone. */
          number: number;
          open_issues: number;
          /**
           * @description The state of the milestone.
           * @enum {string}
           */
          state: "open" | "closed";
          /** @description The title of the milestone. */
          title: string;
          /** Format: date-time */
          updated_at: string;
          /** Format: uri */
          url: string;
        } | null;
        node_id: string;
        number: number;
        /**
         * App
         * @description GitHub apps are a new way to extend GitHub. They can be installed directly on organizations and user accounts and granted access to specific repositories. They come with granular permissions and built-in webhooks. GitHub apps are first class actors within GitHub.
         */
        performed_via_github_app?: {
          /** Format: date-time */
          created_at: string | null;
          description: string | null;
          /** @description The list of events for the GitHub app */
          events?: (
            | "branch_protection_rule"
            | "check_run"
            | "check_suite"
            | "code_scanning_alert"
            | "commit_comment"
            | "content_reference"
            | "create"
            | "delete"
            | "deployment"
            | "deployment_review"
            | "deployment_status"
            | "deploy_key"
            | "discussion"
            | "discussion_comment"
            | "fork"
            | "gollum"
            | "issues"
            | "issue_comment"
            | "label"
            | "member"
            | "membership"
            | "milestone"
            | "organization"
            | "org_block"
            | "page_build"
            | "project"
            | "project_card"
            | "project_column"
            | "public"
            | "pull_request"
            | "pull_request_review"
            | "pull_request_review_comment"
            | "push"
            | "registry_package"
            | "release"
            | "repository"
            | "repository_dispatch"
            | "secret_scanning_alert"
            | "star"
            | "status"
            | "team"
            | "team_add"
            | "watch"
            | "workflow_dispatch"
            | "workflow_run"
            | "reminder"
            | "pull_request_review_thread"
          )[];
          /** Format: uri */
          external_url: string | null;
          /** Format: uri */
          html_url: string;
          /** @description Unique identifier of the GitHub app */
          id: number | null;
          /** @description The name of the GitHub app */
          name: string;
          node_id: string;
          /** User */
          owner: {
            /** Format: uri */
            avatar_url?: string;
            deleted?: boolean;
            email?: string | null;
            /** Format: uri-template */
            events_url?: string;
            /** Format: uri */
            followers_url?: string;
            /** Format: uri-template */
            following_url?: string;
            /** Format: uri-template */
            gists_url?: string;
            gravatar_id?: string;
            /** Format: uri */
            html_url?: string;
            id: number;
            login: string;
            name?: string;
            node_id?: string;
            /** Format: uri */
            organizations_url?: string;
            /** Format: uri */
            received_events_url?: string;
            /** Format: uri */
            repos_url?: string;
            site_admin?: boolean;
            /** Format: uri-template */
            starred_url?: string;
            /** Format: uri */
            subscriptions_url?: string;
            /** @enum {string} */
            type?: "Bot" | "User" | "Organization";
            /** Format: uri */
            url?: string;
          } | null;
          /** @description The set of permissions for the GitHub app */
          permissions?: {
            /** @enum {string} */
            actions?: "read" | "write";
            /** @enum {string} */
            administration?: "read" | "write";
            /** @enum {string} */
            checks?: "read" | "write";
            /** @enum {string} */
            content_references?: "read" | "write";
            /** @enum {string} */
            contents?: "read" | "write";
            /** @enum {string} */
            deployments?: "read" | "write";
            /** @enum {string} */
            discussions?: "read" | "write";
            /** @enum {string} */
            emails?: "read" | "write";
            /** @enum {string} */
            environments?: "read" | "write";
            /** @enum {string} */
            issues?: "read" | "write";
            /** @enum {string} */
            keys?: "read" | "write";
            /** @enum {string} */
            members?: "read" | "write";
            /** @enum {string} */
            metadata?: "read" | "write";
            /** @enum {string} */
            organization_administration?: "read" | "write";
            /** @enum {string} */
            organization_hooks?: "read" | "write";
            /** @enum {string} */
            organization_packages?: "read" | "write";
            /** @enum {string} */
            organization_plan?: "read" | "write";
            /** @enum {string} */
            organization_projects?: "read" | "write" | "admin";
            /** @enum {string} */
            organization_secrets?: "read" | "write";
            /** @enum {string} */
            organization_self_hosted_runners?: "read" | "write";
            /** @enum {string} */
            organization_user_blocking?: "read" | "write";
            /** @enum {string} */
            packages?: "read" | "write";
            /** @enum {string} */
            pages?: "read" | "write";
            /** @enum {string} */
            pull_requests?: "read" | "write";
            /** @enum {string} */
            repository_hooks?: "read" | "write";
            /** @enum {string} */
            repository_projects?: "read" | "write" | "admin";
            /** @enum {string} */
            secret_scanning_alerts?: "read" | "write";
            /** @enum {string} */
            secrets?: "read" | "write";
            /** @enum {string} */
            security_events?: "read" | "write";
            /** @enum {string} */
            security_scanning_alert?: "read" | "write";
            /** @enum {string} */
            single_file?: "read" | "write";
            /** @enum {string} */
            statuses?: "read" | "write";
            /** @enum {string} */
            team_discussions?: "read" | "write";
            /** @enum {string} */
            vulnerability_alerts?: "read" | "write";
            /** @enum {string} */
            workflows?: "read" | "write";
          };
          /** @description The slug name of the GitHub app */
          slug?: string;
          /** Format: date-time */
          updated_at: string | null;
        } | null;
        pull_request?: {
          /** Format: uri */
          diff_url?: string;
          /** Format: uri */
          html_url?: string;
          /** Format: date-time */
          merged_at?: string | null;
          /** Format: uri */
          patch_url?: string;
          /** Format: uri */
          url?: string;
        };
        /** Reactions */
        reactions: {
          "+1": number;
          "-1": number;
          confused: number;
          eyes: number;
          heart: number;
          hooray: number;
          laugh: number;
          rocket: number;
          total_count: number;
          /** Format: uri */
          url: string;
        };
        /** Format: uri */
        repository_url: string;
        /**
         * @description State of the issue; either 'open' or 'closed'
         * @enum {string}
         */
        state?: "open" | "closed";
        state_reason?: string | null;
        /** Format: uri */
        timeline_url?: string;
        /** @description Title of the issue */
        title: string;
        /** Format: date-time */
        updated_at: string;
        /**
         * Format: uri
         * @description URL for the issue
         */
        url: string;
        /** User */
        user: {
          /** Format: uri */
          avatar_url?: string;
          deleted?: boolean;
          email?: string | null;
          /** Format: uri-template */
          events_url?: string;
          /** Format: uri */
          followers_url?: string;
          /** Format: uri-template */
          following_url?: string;
          /** Format: uri-template */
          gists_url?: string;
          gravatar_id?: string;
          /** Format: uri */
          html_url?: string;
          id: number;
          login: string;
          name?: string;
          node_id?: string;
          /** Format: uri */
          organizations_url?: string;
          /** Format: uri */
          received_events_url?: string;
          /** Format: uri */
          repos_url?: string;
          site_admin?: boolean;
          /** Format: uri-template */
          starred_url?: string;
          /** Format: uri */
          subscriptions_url?: string;
          /** @enum {string} */
          type?: "Bot" | "User" | "Organization" | "Mannequin";
          /** Format: uri */
          url?: string;
        } | null;
      } & {
        active_lock_reason?: string | null;
        /** User */
        assignee: {
          /** Format: uri */
          avatar_url?: string;
          deleted?: boolean;
          email?: string | null;
          /** Format: uri-template */
          events_url?: string;
          /** Format: uri */
          followers_url?: string;
          /** Format: uri-template */
          following_url?: string;
          /** Format: uri-template */
          gists_url?: string;
          gravatar_id?: string;
          /** Format: uri */
          html_url?: string;
          id: number;
          login: string;
          name?: string;
          node_id?: string;
          /** Format: uri */
          organizations_url?: string;
          /** Format: uri */
          received_events_url?: string;
          /** Format: uri */
          repos_url?: string;
          site_admin?: boolean;
          /** Format: uri-template */
          starred_url?: string;
          /** Format: uri */
          subscriptions_url?: string;
          /** @enum {string} */
          type?: "Bot" | "User" | "Organization" | "Mannequin";
          /** Format: uri */
          url?: string;
        } | null;
        assignees?: (Record<string, never> | null)[];
        author_association?: string;
        body?: string | null;
        closed_at?: string | null;
        comments?: number;
        comments_url?: string;
        created_at?: string;
        events_url?: string;
        html_url?: string;
        id?: number;
        labels: {
          /** @description 6-character hex code, without the leading #, identifying the color */
          color: string;
          default: boolean;
          description: string | null;
          id: number;
          /** @description The name of the label. */
          name: string;
          node_id: string;
          /**
           * Format: uri
           * @description URL for the label
           */
          url: string;
        }[];
        labels_url?: string;
        locked: boolean;
        milestone?: Record<string, never> | null;
        node_id?: string;
        number?: number;
        performed_via_github_app?: Record<string, never> | null;
        reactions?: {
          "+1"?: number;
          "-1"?: number;
          confused?: number;
          eyes?: number;
          heart?: number;
          hooray?: number;
          laugh?: number;
          rocket?: number;
          total_count?: number;
          url?: string;
        };
        repository_url?: string;
        /**
         * @description State of the issue; either 'open' or 'closed'
         * @enum {string}
         */
        state: "open" | "closed";
        timeline_url?: string;
        title?: string;
        updated_at?: string;
        url?: string;
        user?: {
          avatar_url?: string;
          events_url?: string;
          followers_url?: string;
          following_url?: string;
          gists_url?: string;
          gravatar_id?: string;
          html_url?: string;
          id?: number;
          login?: string;
          node_id?: string;
          organizations_url?: string;
          received_events_url?: string;
          repos_url?: string;
          site_admin?: boolean;
          starred_url?: string;
          subscriptions_url?: string;
          type?: string;
          url?: string;
        };
      };
      organization?: components["schemas"]["organization-simple-webhooks"];
      repository: components["schemas"]["repository-webhooks"];
      sender: components["schemas"]["simple-user-webhooks"];
    };

[0x4007]

My concern is that I feel we should make everything into its own column. However, for simplicity, we could just save the full reaction object in a single column.

[sshivaditya2019]

My concern is that I feel we should make everything into its own column. However, for simplicity, we could just save the full reaction object in a single column.

If we are not going to be using relations and other sql things, we can try with mongoDB or any other document based db ?

[0x4007]

I think Supabase is probably fine because there are no planned features for reactions. We unlikely need the performance for querying for a long time. Even if or when we do, we can consider adding those columns later.

To be honest I'm not a database expert but seems like Supabase should handle our needs. It has a nice UI which is convenient for debugging

[sshivaditya2019]

I think Supabase is probably fine because there are no planned features for reactions. We unlikely need the performance for querying for a long time. Even if or when we do, we can consider adding those columns later.

To be honest I'm not a database expert but seems like Supabase should handle our needs. It has a nice UI which is convenient for debugging

It should be better. I am skeptical about storing serialized comment objects in Postgres database. It should not affect the performance though.

[0x4007]

I asked ChatGPT for a new feature: onboarding new developers, via plaintext q&a. It said we should generate embeddings of an entire repository. How does this affect our schema? Perhaps we should generalize further? Or we can make a new table?

Perhaps a new table per GitHub object type is the most manageable.

[sshivaditya2019]

I asked ChatGPT for a new feature: onboarding new developers, via plaintext q&a. It said we should generate embeddings of an entire repository. How does this affect our schema? Perhaps we should generalize further? Or we can make a new table?

Perhaps a new table per GitHub object type is the most manageable.

that would depend on the onboarding, if onboarding involves explaining the current tickets and comments and the work that's going on, this schema would be enough.

If we want to explain the code and other things, for them we do not need embeddings. For General Q&A this should be enough. We could further expand this by adding OpenAI functions, to query the repo based on the serialized object retrieved using vector search.

[sshivaditya2019]

/start

[ubiquity-os[bot]]

Deadline	Thu, Sep 12, 5:33 PM UTC
Beneficiary	0xDAba6e01D15Db560b88C8F426b016801f79e1F69

Tips:

<ul>
<li>Use <code>/wallet 0x0000...0000</code> if you want to update your registered payment wallet address.</li>
<li>Be sure to open a draft pull request as soon as possible to communicate updates on your progress.</li>
<li>Be sure to provide timely updates to us when requested, or you will be automatically unassigned from the task.</li>
<ul>

[ubiquity-os[bot]]

[ 200 WXDAI ]

@sshivaditya2019

Contributions Overview

View	Contribution	Count	Reward
Issue	Task	1	200
Issue	Comment	18	0
Review	Comment	20	0

Conversation Incentives

Comment	Formatting	Relevance	Reward
@0x4007 would plaintext include issue body as well or would it j…	0 content: p: symbols: \b\w+\b: count: 33 multiplier: 0 score: 1 multiplier: 0	0.7	-
So, plaintext's value would issue body + comment body, or do we …	0 content: p: symbols: \b\w+\b: count: 41 multiplier: 0 score: 1 multiplier: 0	0.8	-
Case 1: Issue Body + Comment Body: - Only one vector index for …	0 content: p: symbols: \b\w+\b: count: 6 multiplier: 0 score: 1 ul: symbols: \b\w+\b: count: 106 multiplier: 0 score: 1 li: symbols: \b\w+\b: count: 35 multiplier: 0 score: 1 code: symbols: \b\w+\b: count: 6 multiplier: 0 score: 1 multiplier: 0	0.8	-
Just to be clear, what is plaintext in your schema ?	0 content: p: symbols: \b\w+\b: count: 10 multiplier: 0 score: 1 multiplier: 0	0.5	-
I'll give an example of plaintext let me know if this right ? &…	0 content: p: symbols: \b\w+\b: count: 27 multiplier: 0 score: 1 pre: symbols: \b\w+\b: count: 62 multiplier: 0 score: 0 code: symbols: \b\w+\b: count: 8 multiplier: 0 score: 1 multiplier: 0	0.7	-
To clarify, the example I shared involves a single embedding/vec…	0 content: p: symbols: \b\w+\b: count: 68 multiplier: 0 score: 1 multiplier: 0	0.6	-
So, to clarify, should the embedding be created per comment whil…	0 content: p: symbols: \b\w+\b: count: 60 multiplier: 0 score: 1 multiplier: 0	0.8	-
An alternative approach would be to store each piece of text fro…	0 content: p: symbols: \b\w+\b: count: 57 multiplier: 0 score: 1 multiplier: 0	0.9	-
To clarify, we don’t provide embeddings directly to the LLM. Ins…	0 content: p: symbols: \b\w+\b: count: 47 multiplier: 0 score: 1 multiplier: 0	0.8	-
@0x4007 Hi, could you update me on the final database schema? …	0 content: p: symbols: \b\w+\b: count: 44 multiplier: 0 score: 1 multiplier: 0	0.7	-
I can probably write a new adapter for them, but I think their f…	0 content: p: symbols: \b\w+\b: count: 32 multiplier: 0 score: 1 multiplier: 0	0.3	-
So, if we add a payment method, the requests per minute can be i…	0 content: p: symbols: \b\w+\b: count: 23 multiplier: 0 score: 1 multiplier: 0	0.4	-
@0x4007 I've added my payment method. Please delete this infor…	0 content: p: symbols: \b\w+\b: count: 24 multiplier: 0 score: 1 multiplier: 0	0.2	-
So, the final schema would be `Id (GlobalNodeId), plaintext(…	0 content: p: symbols: \b\w+\b: count: 14 multiplier: 0 score: 1 code: symbols: \b\w+\b: count: 179 multiplier: 0 score: 1 pre: symbols: \b\w+\b: count: 179 multiplier: 0 score: 0 multiplier: 0	0.9	-
I took this from octokit's type definition. I think reactions sh…	0 content: p: symbols: \b\w+\b: count: 56 multiplier: 0 score: 1 pre: symbols: \b\w+\b: count: 1530 multiplier: 0 score: 0 code: symbols: \b\w+\b: count: 1530 multiplier: 0 score: 1 multiplier: 0	0.8	-
If we are not going to be using relations and other sql things, …	0 content: p: symbols: \b\w+\b: count: 24 multiplier: 0 score: 1 multiplier: 0	0.6	-
It should be better. I am skeptical about storing serialized com…	0 content: p: symbols: \b\w+\b: count: 22 multiplier: 0 score: 1 multiplier: 0	0.5	-
that would depend on the onboarding, if onboarding involves expl…	0 content: p: symbols: \b\w+\b: count: 74 multiplier: 0 score: 1 multiplier: 0	0.7	-
Resolves [#8](https://github.com/ubiquibot/issue-comment-embeddi…	0 content: p: symbols: \b\w+\b: count: 2 multiplier: 0 score: 1 a: symbols: \b\w+\b: count: 1 multiplier: 0 score: 1 ul: symbols: \b\w+\b: count: 40 multiplier: 0 score: 1 li: symbols: \b\w+\b: count: 4 multiplier: 0 score: 1 code: symbols: \b\w+\b: count: 1 multiplier: 0 score: 1 multiplier: 0	1	-
This is according to the schema mentioned in the issue spec, cou…	0 content: p: symbols: \b\w+\b: count: 23 multiplier: 0.2 score: 1 multiplier: 0	1	-
I’ve retained it in case it can be expanded to support OpenAI's …	0 content: p: symbols: \b\w+\b: count: 30 multiplier: 0.2 score: 1 multiplier: 0	1	-
This is the maximum possible length. Even if we switch providers…	0 content: p: symbols: \b\w+\b: count: 33 multiplier: 0.2 score: 1 multiplier: 0	1	-
I wanted to keep it consistent with the DB schema. This passed t…	0 content: p: symbols: \b\w+\b: count: 24 multiplier: 0.2 score: 1 multiplier: 0	1	-
Its there to ignore, bot comments, and chore issues created. So,…	0 content: p: symbols: \b\w+\b: count: 38 multiplier: 0.2 score: 1 multiplier: 0	1	-
I don't think its safe to give codeblock to jsdom. This works fi…	0 content: p: symbols: \b\w+\b: count: 31 multiplier: 0.2 score: 1 multiplier: 0	1	-
Replace it with `Markdown-it`. This library is being mai…	0 content: p: symbols: \b\w+\b: count: 15 multiplier: 0.2 score: 1 code: symbols: \b\w+\b: count: 2 multiplier: 0.2 score: 1 multiplier: 0	1	-
If a text is empty, instead of calling embedding api on it I am …	0 content: p: symbols: \b\w+\b: count: 56 multiplier: 0.2 score: 1 multiplier: 0	1	-
This can be used for issue deduplication and stuff. I think this…	0 content: p: symbols: \b\w+\b: count: 16 multiplier: 0.2 score: 1 multiplier: 0	1	-
Fixed	0 content: p: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 1 multiplier: 0	1	-
@0x4007 Updated the schema, comments use `voyageai` for …	0 content: p: symbols: \b\w+\b: count: 10 multiplier: 0.2 score: 1 code: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 1 multiplier: 0	1	-
@0x4007 Could you please check the updated changes ? Have remove…	0 content: p: symbols: \b\w+\b: count: 23 multiplier: 0.2 score: 1 code: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 1 multiplier: 0	1	-
I have removed the changes for the other task. CI Should be pass…	0 content: p: symbols: \b\w+\b: count: 25 multiplier: 0.2 score: 1 multiplier: 0	1	-
Right now issue body is in the payload object, this is jsonb, wh…	0 content: p: symbols: \b\w+\b: count: 18 multiplier: 0.2 score: 1 img: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 0 multiplier: 0	1	-
So to sum up, you require two tables one for `comments` …	0 content: p: symbols: \b\w+\b: count: 44 multiplier: 0.2 score: 1 code: symbols: \b\w+\b: count: 7 multiplier: 0.2 score: 1 multiplier: 0	1	-
@0x4007 Have Added Two Separate Tables as per the schema mention…	0 content: p: symbols: \b\w+\b: count: 26 multiplier: 0.2 score: 1 multiplier: 0	1	-
Yes, it includes a payload. I've retained the type column to dis…	0 content: p: symbols: \b\w+\b: count: 38 multiplier: 0.2 score: 1 multiplier: 0	1	-
Removed the `type` from schema. Payload is stored for bo…	0 content: p: symbols: \b\w+\b: count: 24 multiplier: 0.2 score: 1 code: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 1 multiplier: 0	1	-
@0x4007 I have added `markdown` and `plaintext` …	0 content: p: symbols: \b\w+\b: count: 39 multiplier: 0.2 score: 1 code: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 1 multiplier: 0	1	-

[ 114.801 WXDAI ]

@0x4007

Contributions Overview

View	Contribution	Count	Reward
Issue	Specification	1	16.82
Issue	Comment	15	59.741
Review	Comment	27	38.24

Conversation Incentives

Comment	Formatting	Relevance	Reward
This requires changes to the plugin that captures the data, as w…	16.82 content: h2: symbols: \b\w+\b: count: 21 multiplier: 0.1 score: 1 img: symbols: \b\w+\b: count: 1 multiplier: 0.1 score: 5 p: symbols: \b\w+\b: count: 19 multiplier: 0.1 score: 1 pre: symbols: \b\w+\b: count: 6 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 1 multiplier: 0.1 score: 5 ul: symbols: \b\w+\b: count: 192 multiplier: 0.1 score: 0 li: symbols: \b\w+\b: count: 35 multiplier: 0.1 score: 1 multiplier: 3	1	16.82
Issue body as well and no leave null	1.17 content: p: symbols: \b\w+\b: count: 8 multiplier: 0.2 score: 1 multiplier: 1	0.5	0.585
Why do you want to associate and/or combine the embedding of the…	3.6 content: p: symbols: \b\w+\b: count: 30 multiplier: 0.2 score: 1 multiplier: 1	0.7	2.52
I don't understand your explanation. Anyways would be great if …	8.07 content: p: symbols: \b\w+\b: count: 17 multiplier: 0.2 score: 1 ol: symbols: \b\w+\b: count: 55 multiplier: 0.2 score: 0 li: symbols: \b\w+\b: count: 53 multiplier: 0.2 score: 1 multiplier: 1	0.8	6.456
The source code of the markdown of the comments. I hope that th…	7.67 content: p: symbols: \b\w+\b: count: 73 multiplier: 0.2 score: 1 multiplier: 1	0.4	3.068
I am not an expert working with embeddings but as I understand w…	4.21 content: p: symbols: \b\w+\b: count: 36 multiplier: 0.2 score: 1 multiplier: 1	0.6	2.526
So then if we are clobbering a single embedding to represent an …	4.99 content: p: symbols: \b\w+\b: count: 44 multiplier: 0.2 score: 1 multiplier: 1	0.7	3.493
@gentlementlegen im assuming storing node id won't be sufficient…	7.04 content: p: symbols: \b\w+\b: count: 66 multiplier: 0.2 score: 1 multiplier: 1	0.8	5.632
OpenAI’s text embeddings measure the relatedness of text strings…	12.3 content: h2: symbols: \b\w+\b: count: 8 multiplier: 0.2 score: 1 p: symbols: \b\w+\b: count: 103 multiplier: 0.2 score: 1 a: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 1 code: symbols: \b\w+\b: count: 4 multiplier: 0.2 score: 1 multiplier: 1	0.9	11.07
Three requests per minute seems acceptable as long as we can hav…	10.79 content: p: symbols: \b\w+\b: count: 109 multiplier: 0.2 score: 1 multiplier: 1	0.8	8.632
![image](https://github.com/user-attachments/assets/eaf5c732-5c8…	1.05 content: p: symbols: \b\w+\b: count: 7 multiplier: 0.2 score: 1 img: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 0 multiplier: 1	0.1	0.105
It's merchant locked for up to $1	1.17 content: p: symbols: \b\w+\b: count: 8 multiplier: 0.2 score: 1 multiplier: 1	0.1	0.117
Looks like there isn't a lot to save from there if we remove the…	6.95 content: p: symbols: \b\w+\b: count: 65 multiplier: 0.2 score: 1 br: symbols: \b\w+\b: count: 1 multiplier: 0.2 score: 0 multiplier: 1	0.7	4.865
My concern is that I feel we should make everything into its own…	3.5 content: p: symbols: \b\w+\b: count: 29 multiplier: 0.2 score: 1 multiplier: 1	0.5	1.75
I think Supabase is probably fine because there are no planned f…	6.95 content: p: symbols: \b\w+\b: count: 65 multiplier: 0.2 score: 1 multiplier: 1	0.6	4.17
I asked ChatGPT for a new feature: onboarding new developers, vi…	5.94 content: p: symbols: \b\w+\b: count: 54 multiplier: 0.2 score: 1 multiplier: 1	0.8	4.752
Seems generally okay. Let me see how your database looks	0.71 content: p: symbols: \b\w+\b: count: 10 multiplier: 0.1 score: 1 multiplier: 1	1	0.71
```suggestion const authorId = payload.comment.use…	0.52 content: pre: symbols: \b\w+\b: count: 7 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 7 multiplier: 0.1 score: 1 multiplier: 1	1	0.52
Sync with the GitHub metadata of the comment or issue	0.71 content: p: symbols: \b\w+\b: count: 10 multiplier: 0.1 score: 1 multiplier: 1	1	0.71
```suggestion ```	0.1 content: pre: symbols: \b\w+\b: count: 1 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 1 multiplier: 0.1 score: 1 multiplier: 1	1	0.1
```suggestion plaintext = null; ``&#…	0.64 content: pre: symbols: \b\w+\b: count: 2 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 2 multiplier: 0.1 score: 1 p: symbols: \b\w+\b: count: 6 multiplier: 0.1 score: 1 multiplier: 1	1	0.64
```suggestion plaintext = null; ```	0.18 content: pre: symbols: \b\w+\b: count: 2 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 2 multiplier: 0.1 score: 1 multiplier: 1	1	0.18
Should this file be deleted or something?	0.52 content: p: symbols: \b\w+\b: count: 7 multiplier: 0.1 score: 1 multiplier: 1	1	0.52
Why does it have to be this exact length?	0.65 content: p: symbols: \b\w+\b: count: 9 multiplier: 0.1 score: 1 multiplier: 1	1	0.65
```suggestion commentObject?: Record<string, un…	1.5 content: pre: symbols: \b\w+\b: count: 5 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 5 multiplier: 0.1 score: 1 p: symbols: \b\w+\b: count: 17 multiplier: 0.1 score: 1 multiplier: 1	1	1.5
Seems unnecessary. Just delete the trailing zeros.	0.52 content: p: symbols: \b\w+\b: count: 7 multiplier: 0.1 score: 1 multiplier: 1	1	0.52
We can restore from git history if needed. Its not in use so it …	1.11 content: p: symbols: \b\w+\b: count: 17 multiplier: 0.1 score: 1 multiplier: 1	1	1.11
If you defined it, then its wrong. If its from another database …	1.54 content: p: symbols: \b\w+\b: count: 25 multiplier: 0.1 score: 1 multiplier: 1	1	1.54
Why is the default issue on the comments table? Why even add the…	1.38 content: p: symbols: \b\w+\b: count: 22 multiplier: 0.1 score: 1 multiplier: 1	1	1.38
This seems very error prone. Why dont you use some virtual DOM (…	1.46 content: p: symbols: \b\w+\b: count: 20 multiplier: 0.1 score: 1 code: symbols: \b\w+\b: count: 2 multiplier: 0.1 score: 1 multiplier: 1	1	1.46
```suggestion model: "voyage-large-2-instru…	1.02 content: pre: symbols: \b\w+\b: count: 5 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 5 multiplier: 0.1 score: 1 p: symbols: \b\w+\b: count: 7 multiplier: 0.1 score: 1 a: symbols: \b\w+\b: count: 1 multiplier: 0.1 score: 1 multiplier: 1	1	1.02
```suggestion const markdown = payload.issue.body…	1.92 content: pre: symbols: \b\w+\b: count: 9 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 9 multiplier: 0.1 score: 1 p: symbols: \b\w+\b: count: 20 multiplier: 0.1 score: 1 multiplier: 1	1	1.92
```suggestion logger.debug(`Exiting addIssue&…	1.71 content: pre: symbols: \b\w+\b: count: 4 multiplier: 0.1 score: 0 code: symbols: \b\w+\b: count: 4 multiplier: 0.1 score: 1 p: symbols: \b\w+\b: count: 22 multiplier: 0.1 score: 1 multiplier: 1	1	1.71
This seems out of scope?	0.39 content: p: symbols: \b\w+\b: count: 5 multiplier: 0.1 score: 1 multiplier: 1	1	0.39
Still confused about this fill stuff	0.46 content: p: symbols: \b\w+\b: count: 6 multiplier: 0.1 score: 1 multiplier: 1	1	0.46
Why did you switch to main from development that doesn't seem ri…	0.88 content: p: symbols: \b\w+\b: count: 13 multiplier: 0.1 score: 1 multiplier: 1	1	0.88
You'll need to cherry pick changes (easy to do with a git UI) an…	4.58 content: p: symbols: \b\w+\b: count: 90 multiplier: 0.1 score: 1 multiplier: 1	1	4.58
Where's the issue body? You should probably make another table a…	1.75 content: p: symbols: \b\w+\b: count: 29 multiplier: 0.1 score: 1 multiplier: 1	1	1.75
I'm bad at deciding this sort of thing. Let's go with your sugge…	1.7 content: p: symbols: \b\w+\b: count: 28 multiplier: 0.1 score: 1 multiplier: 1	1	1.7
Seems mostly good but I didn't see all the headers on the first …	2.69 content: p: symbols: \b\w+\b: count: 48 multiplier: 0.1 score: 1 multiplier: 1	1	2.69
Yes I think its unnecessary if they are separated by type on dif…	0.94 content: p: symbols: \b\w+\b: count: 14 multiplier: 0.1 score: 1 multiplier: 1	1	0.94
Thanks for the thorough QA. You don't need to make a new video o…	7.49 content: p: symbols: \b\w+\b: count: 158 multiplier: 0.1 score: 1 code: symbols: \b\w+\b: count: 1 multiplier: 0.1 score: 1 multiplier: 1	1	7.49
I think that only tier5 subscribers can use right now via API. I…	1.17 content: p: symbols: \b\w+\b: count: 18 multiplier: 0.1 score: 1 multiplier: 1	1	1.17

[0x4007]

I'll need to top up the wallet soon