my prefect server setup
prefect-metrics.waow.tech
python
orchestration
grafana: fix DuckDB datasource and deploy github intelligence dashboard
- remove readOnly:true from analytics volume mount — DuckDB needs
filesystem write access for lock files even in read mode
- update github-intelligence dashboard table panel to query
github_action_items (dbt-scored) instead of raw inline formula
- deploy dashboard via just dashboards
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
+3
-4
+3
-3
deploy/dashboards/github-intelligence.json
+3
-3
deploy/dashboards/github-intelligence.json
···
61
61
"id": 5,
62
62
"type": "table",
63
63
"title": "top action items (open, scored by engagement)",
64
-
"description": "recency-weighted by updated_at, boosted by reactions + comments. once dbt enrich flow has run, query github_action_items instead.",
64
+
"description": "scored by dbt: recency × engagement × label multiplier × contributor weight",
65
65
"gridPos": {"h": 12, "w": 24, "x": 0, "y": 4},
66
66
"datasource": {"type": "motherduck-duckdb-datasource", "uid": "duckdb"},
67
-
"options": {"sortBy": [{"desc": true, "displayName": "score"}]},
67
+
"options": {"sortBy": [{"desc": true, "displayName": "importance_score"}]},
68
68
"targets": [
69
69
{
70
-
"rawSql": "SELECT\n repo,\n number,\n type,\n title,\n url,\n \"user\",\n comments,\n reactions_total,\n ROUND(\n GREATEST(0.0, 1.0 - DATEDIFF('day', updated_at::DATE, CURRENT_DATE) / 30.0)\n * LEAST(1.0, LN(1 + comments + reactions_total) / LN(50))\n , 4) AS score,\n updated_at\nFROM raw_github_issues\nWHERE state = 'open'\nORDER BY score DESC, comments DESC\nLIMIT 50",
70
+
"rawSql": "SELECT repo, number, type, title, url, \"user\", labels, comments, reactions_total, importance_score, updated_at FROM github_action_items ORDER BY importance_score DESC",
71
71
"format": "table"
72
72
}
73
73
],