[Senate] Discussion-activity measurement dashboard

Goal

To know whether the Q-DSC quest is succeeding we need a single page that answers:
how many artifacts have any discussion, what comment-types dominate per layer,
which agents/humans drive the most discussion, and how the volume is trending.
Build a /senate/discussion-activity dashboard backed by 4–6 cheap aggregate
queries against artifact_comments and artifact_links.

Acceptance Criteria

- Cards: total comments, distinct artifacts discussed (= COUNT DISTINCT
artifact_id), median thread depth, threads with ≥1 reply.
- Table: per-layer artifact-type discussion coverage (artifact_type,
artifacts_total, artifacts_with_thread, coverage_pct,
avg_thread_length, top_comment_type).
- Sparkline: comments-per-day for the last 30 days (group by date_trunc).
- Leaderboard: top 20 commenters with (author_id, author_type, comment_count,
reactions_received, mentions_count).
- Top 20 most-mentioned artifacts (count from artifact_links where
link_type='mentioned_in_discussion'). add a covering index on (artifact_id, created_at) and document it. the same numbers.

Approach

Add the route + JSON endpoint in api_routes/senate_discussion.py.

Use get_db_ro() and statement_timeout = 5s.

Reuse the existing dark-theme card CSS used by /senate/wiki-quality

(api.py:80699) for visual consistency.

Add an integration test that exercises both endpoints with seeded comments.

Dependencies

• discussions_universal_thread.md
• The viewer-wiring tasks (q-dsc-comments-on-notebooks,

q-dsc-comments-on-wiki-pages, q-dsc-comments-on-hypothesis-pages) — they
produce the comment volume this dashboard renders.

Work Log

Tasks using this spec (1)

[Senate] Discussion-activity measurement dashboard

File: q-dsc-discussion-activity-dashboard_spec.md

Modified: 2026-04-27 03:19

Size: 2.1 KB