Add nightly Moonbeam argument leaderboard extraction, API, and muzzle.lol review UI by Copilot · Pull Request #258 · dev-chat/mocker

Copilot · 2026-05-21T14:42:33Z

Identify the single review comment to address and inspect the affected backend code
Run backend lint, build, and tests to establish the current baseline
Make the smallest change needed to limit or paginate argument history loading
Run targeted validation for the touched backend code
Run final review validation and report the completed fix

Original prompt

DB Layer:

Create a new table for argument leaderboards capturing:

Argument summary and each participant's side/viewpoint (text/JSON field).

Winner of the argument (foreign key to slack_user).

Point value (integer, 0–5), representing the substance and depth of the argument.

Service Layer:

When a user asks "who won an argument" in Moonbeam:

Use ChatGPT to determine:

The argument context.

Each participant and their viewpoints.

The winner (as mapped to a slack_user DB record).

Calculate point valuation (0–5) based on argument substance (length, back-and-forth, message count, etc.).

Save extracted data (argument details, winner, points) in the new DB table.

UI Layer:

Add a page to the muzzle.lol website displaying:

The argument leaderboard (ranking users by argument wins/points).

Listed columns: argument, each participant's side, winner, point value.

An intuitive leaderboard at the top (argument wins by user).

Enable navigation to review detailed outcomes of previous arguments.

Ensure the UI is clear, readable, and user-friendly, making it easy to identify top debaters and review past debates.

This pull request was created from Copilot chat.

Agent-Logs-Url: https://github.com/dev-chat/mocker/sessions/92100499-66e6-4c96-b4c3-e1e87c3ebf96

sfreeman422 · 2026-05-21T15:13:07Z

@copilot refactor this to instead behave in such a way that as part of it's nightly job, this service will analyze all channels for arguments between users and determine who won. it should be intolerant of false positives, IE very selective in the conversations it believes are arguments. We should not be weakly checking if a tagged message is asking who won etc. This should largely be handled by the AI

Copilot · 2026-05-21T15:21:31Z

@copilot refactor this to instead behave in such a way that as part of it's nightly job, this service will analyze all channels for arguments between users and determine who won. it should be intolera...

Addressed in a9a2285. Argument extraction now runs in a selective nightly analysis job across channels, and the tagged-message winner check was removed from Moonbeam participation.

sfreeman422 · 2026-05-21T17:16:04Z

+- any conversation where the winner is ambiguous
+- weak or low-substance disagreements that do not merit leaderboard tracking
+
+If you find a qualifying argument, return ONLY a single JSON object in this exact shape:


@copilot this should be an array.

Fixed in 6d22437. The extractor prompt and parser now use a JSON array response shape ([] when nothing qualifies).

sfreeman422 · 2026-05-21T17:49:39Z

@copilot fix the failing checks come on COME ON

Copilot · 2026-05-21T17:56:41Z

@copilot fix the failing checks come on COME ON

Fixed in f91b236. I traced the failing CI job to the backend coverage gate (79.8% branch coverage) and added focused dashboard controller tests to cover the new authenticated dashboard routes, including /dashboard/arguments.

Copilot

Pull request overview

Copilot reviewed 26 out of 26 changed files in this pull request and generated 3 comments.

Comments suppressed due to low confidence (2)

packages/backend/src/argument/argument.persistence.service.ts:88

Participants are resolved without filtering out bot SlackUser records. That can allow bot accounts to be stored in participants / participantViewpoints and appear in the UI. Consider adding isBot: false to the participant lookup (and keeping the "at least 2 participants" check as "at least 2 human participants").

    const participantUsers = await slackUserRepo.find({
      where: participants.map((participant) => ({ slackId: participant.slackId, teamId: input.teamId })),
    });

packages/backend/src/argument/argument.persistence.service.ts:156

The leaderboard SQL does not filter out bot winners. Other leaderboards in the codebase exclude bots (e.g. u.isBot = 0 in DashboardPersistenceService), so adding the same filter here will prevent bot users from ranking on the argument leaderboard.

          `SELECT u.name AS name,
                  u.slackId AS slackId,
                  CAST(COUNT(*) AS SIGNED) AS wins,
                  CAST(COALESCE(SUM(a.pointValue), 0) AS SIGNED) AS points
            FROM argument_leaderboard a
            INNER JOIN slack_user u ON u.id = a.winnerId
            WHERE a.teamId = ?
            GROUP BY u.id, u.name, u.slackId
            ORDER BY wins DESC, points DESC, u.name ASC`,

Copilot · 2026-05-22T16:43:29Z

+    const slackUserRepo = getRepository(SlackUser);
+    const winner = await slackUserRepo.findOne({
+      where: { slackId: input.winnerSlackId, teamId: input.teamId },
+    });


Fixed in 40a5c51. Winners and participant lookups now require isBot: false, and the leaderboard aggregate query also excludes bot winners.

Copilot · 2026-05-22T16:42:07Z

+    const participantSlackIds = extractParticipantSlackIds(historyMessages, {
+      excludeSlackIds: [MOONBEAM_SLACK_ID],
+    });
+    if (participantSlackIds.length < 2) return;
+
+    await this.extractArgument(teamId, channelId, historyMessages);


Fixed in ce9c4b4. The job now checks the extracted participant Slack IDs against known non-bot SlackUser records and skips the OpenAI call unless at least two human participants remain.

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Copilot

Pull request overview

Copilot reviewed 26 out of 26 changed files in this pull request and generated 4 comments.

Comments suppressed due to low confidence (4)

packages/backend/src/argument/argument.persistence.service.ts:92

slackUserRepo.find({ where: participants.map(...) }) can be called with participants.length === 0, which (depending on TypeORM behavior) may result in an unfiltered query. Add an early return when fewer than 2 participants remain after normalization, and include isBot: false in the participant lookup to prevent bot records from being attached.

    const participants = Array.from(
      new Map(
        input.participants
          .map(normalizeParticipant)
          .filter((participant): participant is ArgumentParticipant => participant !== null)
          .map((participant) => [participant.slackId, participant]),
      ).values(),
    );

    const participantUsers = await slackUserRepo.find({
      where: participants.map((participant) => ({ slackId: participant.slackId, teamId: input.teamId, isBot: false })),
    });
    const extractedParticipantViewpoints = buildParticipantViewpoints(participants);
    const resolvedParticipants = buildArgumentParticipants(participantUsers, extractedParticipantViewpoints);
    const participantViewpoints = buildParticipantViewpoints(resolvedParticipants);

packages/backend/src/argument/argument.persistence.service.ts:99

After resolving participants, there is no validation that the resolved participant set still contains winnerSlackId. If the winner exists but is not among the resolved participants (e.g., participant resolution failed), you can persist an outcome where the winner wasn’t a participant. Consider skipping the save unless the winner is included in the resolved participants list.

    const extractedParticipantViewpoints = buildParticipantViewpoints(participants);
    const resolvedParticipants = buildArgumentParticipants(participantUsers, extractedParticipantViewpoints);
    const participantViewpoints = buildParticipantViewpoints(resolvedParticipants);

    if (resolvedParticipants.length < 2) {
      this.logger.warn('Skipping argument outcome save because fewer than two participants were extracted', {
        teamId: input.teamId,
        channelId: input.channelId,
      });
      return null;
    }

packages/backend/src/argument/argument.persistence.service.ts:156

The aggregate leaderboard SQL does not exclude bot winners (missing u.isBot = 0). If a bot is ever stored as a winner, it will be ranked here. Add the same bot filter used elsewhere in dashboard queries.

        repo.query<LeaderboardRow[]>(
          `SELECT u.name AS name,
                  u.slackId AS slackId,
                  CAST(COUNT(*) AS SIGNED) AS wins,
                  CAST(COALESCE(SUM(a.pointValue), 0) AS SIGNED) AS points
            FROM argument_leaderboard a
            INNER JOIN slack_user u ON u.id = a.winnerId
            WHERE a.teamId = ? AND u.isBot = 0
             GROUP BY u.id, u.name, u.slackId
            ORDER BY wins DESC, points DESC, u.name ASC`,

packages/backend/src/argument/argument.persistence.service.spec.ts:174

The getArgumentLeaderboard test only asserts the COUNT cast is present in the SQL. Once the query is updated to exclude bots, add an assertion that the SQL contains u.isBot = 0 so bot filtering can’t regress silently.

        winner: { id: 2, slackId: 'U2', name: 'Bob' },
        pointValue: '4',
        createdAt: '2026-05-21T00:00:00.000Z',
      },

+    const slackUserRepo = getRepository(SlackUser);
+    const winner = await slackUserRepo.findOne({
+      where: { slackId: input.winnerSlackId, teamId: input.teamId, isBot: false },
+    });


+        repo.find({
+          where: { teamId },
+          relations: ['participants', 'winner'],
+          order: { createdAt: 'DESC' },
+        }),


+  it('saves an argument outcome with a resolved winner', async () => {
+    findOne.mockResolvedValue({ id: 7, slackId: 'U2', name: 'Bob' });
+    findUsers.mockResolvedValue([
+      { id: 6, slackId: 'U1', name: 'Alice' },
+      { id: 7, slackId: 'U2', name: 'Bob' },
+    ]);
+    save.mockImplementation(async (entry: { createdAt?: Date }) => ({
+      ...entry,
+      id: 11,
+      createdAt: entry.createdAt ?? new Date('2026-05-21T00:00:00.000Z'),
+    }));
+
+    const result = await service.saveArgumentOutcome({
+      teamId: 'T1',
+      channelId: 'C1',
+      argumentSummary: 'Tabs versus spaces',
+      participants: [
+        { slackId: 'U1', name: 'Alice', viewpoint: 'tabs are faster' },
+        { slackId: 'U2', name: 'Bob', viewpoint: 'spaces are clearer' },
+      ],
+      winnerSlackId: 'U2',
+      pointValue: 6,
+    });
+
+    expect(findOne).toHaveBeenCalledWith({ where: { slackId: 'U2', teamId: 'T1', isBot: false } });
+    expect(findUsers).toHaveBeenCalledWith({
+      where: [


Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Copilot

Pull request overview

Copilot reviewed 26 out of 26 changed files in this pull request and generated 2 comments.

+        repo.find({
+          where: { teamId },
+          relations: ['participants', 'winner'],
+          order: { createdAt: 'DESC' },
+          take: ARGUMENT_HISTORY_LIMIT,
+        }),


+    const participantViewpoints = buildParticipantViewpoints(resolvedParticipants);
+


Initial plan

fc70994

Copilot AI assigned Copilot and sfreeman422 May 21, 2026

Copilot started work on behalf of sfreeman422 May 21, 2026 14:42 View session

feat: add argument leaderboard page

34efe91

Agent-Logs-Url: https://github.com/dev-chat/mocker/sessions/92100499-66e6-4c96-b4c3-e1e87c3ebf96

Copilot AI changed the title ~~[WIP] Add new table for argument leaderboards~~ Add Moonbeam argument leaderboard persistence, API, and muzzle.lol review UI May 21, 2026

Copilot AI requested a review from sfreeman422 May 21, 2026 15:01

Copilot finished work on behalf of sfreeman422 May 21, 2026 15:01

Copilot started work on behalf of sfreeman422 May 21, 2026 15:13 View session

refactor: move argument extraction to nightly job

a9a2285

Copilot AI changed the title ~~Add Moonbeam argument leaderboard persistence, API, and muzzle.lol review UI~~ Add nightly Moonbeam argument leaderboard extraction, API, and muzzle.lol review UI May 21, 2026

Copilot finished work on behalf of sfreeman422 May 21, 2026 15:23

sfreeman422 requested changes May 21, 2026

View reviewed changes

Copilot started work on behalf of sfreeman422 May 21, 2026 17:18 View session

fix: support multiple nightly argument extractions

6d22437

Copilot finished work on behalf of sfreeman422 May 21, 2026 17:25

Copilot AI requested a review from sfreeman422 May 21, 2026 17:25

Copilot started work on behalf of sfreeman422 May 21, 2026 17:27 View session

Copilot AI added 3 commits May 21, 2026 17:32

fix: store argument participants as slack users

2ad28e5

fix: clean up argument participant persistence

b11a8f2

fix: tidy argument participant validation

1f29b37

Copilot finished work on behalf of sfreeman422 May 21, 2026 17:42

Copilot started work on behalf of sfreeman422 May 21, 2026 17:49 View session

Add dashboard controller coverage

f91b236

Copilot finished work on behalf of sfreeman422 May 21, 2026 17:57

sfreeman422 marked this pull request as ready for review May 21, 2026 18:25

fix: keep argument participant viewpoints in sync

58cadab

Copilot finished work on behalf of sfreeman422 May 22, 2026 15:07

Copilot started work on behalf of sfreeman422 May 22, 2026 15:14 View session

Copilot finished work on behalf of sfreeman422 May 22, 2026 15:15

sfreeman422 requested a review from Copilot May 22, 2026 15:18

Copilot started reviewing on behalf of sfreeman422 May 22, 2026 15:19 View session

Copilot AI reviewed May 22, 2026

View reviewed changes

Potential fix for pull request finding

97b5a7b

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Copilot started work on behalf of sfreeman422 May 22, 2026 16:37 View session

Copilot AI added 2 commits May 22, 2026 16:40

fix: skip argument extraction when bots leave fewer than two humans

ce9c4b4

fix: exclude bot users from argument winners

40a5c51

Copilot finished work on behalf of sfreeman422 May 22, 2026 16:42

sfreeman422 requested a review from Copilot May 22, 2026 16:42

Copilot started reviewing on behalf of sfreeman422 May 22, 2026 16:43 View session

Copilot finished work on behalf of sfreeman422 May 22, 2026 16:44

Copilot AI reviewed May 22, 2026

View reviewed changes

Copilot started work on behalf of sfreeman422 May 22, 2026 16:54 View session

Copilot started work on behalf of sfreeman422 May 22, 2026 16:55 View session

sfreeman422 and others added 2 commits May 22, 2026 12:55

Potential fix for pull request finding

cb3bc57

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

test: cover bot exclusion in argument persistence specs

17c7c52

Copilot finished work on behalf of sfreeman422 May 22, 2026 16:58

fix: cap argument history query

a04b28c

Copilot finished work on behalf of sfreeman422 May 22, 2026 16:59

Copilot finished work on behalf of sfreeman422 May 22, 2026 17:01

sfreeman422 requested a review from Copilot May 22, 2026 17:03

Copilot started reviewing on behalf of sfreeman422 May 22, 2026 17:04 View session

Copilot AI reviewed May 22, 2026

View reviewed changes

Merge branch 'master' into copilot/add-argument-leaderboard-table

646c676

		const participantViewpoints = buildParticipantViewpoints(resolvedParticipants);

Conversation

Copilot AI commented May 21, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

sfreeman422 commented May 21, 2026

Uh oh!

Copilot AI commented May 21, 2026

Uh oh!

Uh oh!

sfreeman422 May 21, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI May 21, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

sfreeman422 commented May 21, 2026

Uh oh!

Copilot AI commented May 21, 2026

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI May 22, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI May 22, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Copilot AI commented May 21, 2026 •

edited

Loading