cw-wrapped

name	cw:wrapped
description	Generate a fun, shareable summary of your Cursor agent usage stats — like Spotify Wrapped but for AI-assisted development. Use when the user asks for their stats, summary, wrapped, or wants a fun overview of their Cursor activity.

Wrapped — Your Cursor Year in Review

Generate a fun, visual summary of the user's Cursor agent activity. Think Spotify Wrapped energy.

Run ALL queries below, then present as an engaging, shareable summary.

Finding the script

CURSOR_PLUGIN_ROOT should be set when invoked through the plugin system, but may be unset during development. Resolve once per session:

Use find -L so symlinked dev installs (e.g. local/cursor-warehouse → your clone) are traversed.

QUERY_SCRIPT="${CURSOR_PLUGIN_ROOT:+$CURSOR_PLUGIN_ROOT/scripts/query.py}"
if [ -z "$QUERY_SCRIPT" ] || [ ! -f "$QUERY_SCRIPT" ]; then
  QUERY_SCRIPT="$(find -L ~/.cursor/plugins -name query.py -path '*/cursor-warehouse/*/query.py' 2>/dev/null | head -1)"
fi

Then use uv run --script "$QUERY_SCRIPT" sql "..." for all queries below.

Schema reference (for adapting queries)

When the user specifies a date range, add WHERE clauses. Use these join paths:

sessions.session_id  ←→  messages.session_id
sessions.session_id  ←→  tool_calls.session_id
messages.uuid        ←→  tool_calls.message_uuid   (NOT message_id)

To filter tool_calls or messages by date, JOIN through sessions:

SELECT tc.tool_name, COUNT(*) uses
FROM tool_calls tc
JOIN sessions s ON tc.session_id = s.session_id
WHERE s.created_at >= '...' AND s.created_at < '...'
GROUP BY 1 ORDER BY 2 DESC

SELECT model, COUNT(*) messages
FROM messages m
JOIN sessions s ON m.session_id = s.session_id
WHERE model IS NOT NULL AND s.created_at >= '...' AND s.created_at < '...'
GROUP BY 1 ORDER BY 2 DESC

Data collection

All-time stats

uv run --script "$QUERY_SCRIPT" sql "SELECT COUNT(*) total_sessions, SUM(message_count) total_messages, COUNT(DISTINCT project_name) total_projects, MIN(created_at)::DATE first_session, MAX(created_at)::DATE latest_session FROM sessions"

Top projects by session count

uv run --script "$QUERY_SCRIPT" sql "SELECT project_name, COUNT(*) sessions, SUM(message_count) messages FROM sessions GROUP BY 1 ORDER BY 2 DESC LIMIT 5"

Favorite tools (top 10)

uv run --script "$QUERY_SCRIPT" sql "SELECT tc.tool_name, COUNT(*) uses FROM tool_calls tc GROUP BY 1 ORDER BY 2 DESC LIMIT 10"

Longest session ever

NOTE: first_prompt contains raw Cursor system context (XML tags). Use regexp_extract to get the actual user prompt:

uv run --script "$QUERY_SCRIPT" sql "SELECT project_name, created_at::DATE, message_count, LEFT(COALESCE(NULLIF(regexp_extract(first_prompt, '<user_query>\s*([\s\S]*?)\s*</user_query>', 1), ''), NULLIF(regexp_extract(first_prompt, '(/\S+[^\n<]*)', 1), ''), first_prompt), 200) AS prompt FROM sessions ORDER BY message_count DESC LIMIT 1"

Busiest day

uv run --script "$QUERY_SCRIPT" sql "SELECT created_at::DATE as day, COUNT(*) sessions, SUM(message_count) messages FROM sessions GROUP BY 1 ORDER BY 2 DESC LIMIT 1"

Models used

uv run --script "$QUERY_SCRIPT" sql "SELECT model, COUNT(*) messages FROM messages WHERE model IS NOT NULL GROUP BY 1 ORDER BY 2 DESC LIMIT 5"

Streak (consecutive days)

uv run --script "$QUERY_SCRIPT" sql "WITH days AS (SELECT DISTINCT created_at::DATE as d FROM sessions), streaks AS (SELECT d, d - ROW_NUMBER() OVER (ORDER BY d) * INTERVAL '1 day' as grp FROM days) SELECT COUNT(*) as streak_days, MIN(d)::DATE as from_date, MAX(d)::DATE as to_date FROM streaks GROUP BY grp ORDER BY streak_days DESC LIMIT 1"

Session distribution by hour of day

uv run --script "$QUERY_SCRIPT" sql "SELECT EXTRACT(HOUR FROM created_at) as hour, COUNT(*) sessions FROM sessions GROUP BY 1 ORDER BY 2 DESC LIMIT 3"

AI attribution (if available)

uv run --script "$QUERY_SCRIPT" sql "SELECT COUNT(*) total_commits, SUM(lines_added) total_lines, AVG(CAST(REPLACE(COALESCE(v2_ai_percentage, '0'), '%', '') AS FLOAT)) avg_ai_pct FROM scored_commits"

Presentation

Present as a fun, engaging summary with personality. Use section headers like:

• Your Numbers — total sessions, messages, projects
• #1 Project — your most-visited project and what it says about you
• Power Tools — your top 5 tools and what that means
• Marathon Session — your longest session: what happened?
• Peak Hours — when you do your best AI-assisted work
• Your Streak — longest consecutive days using Cursor
• AI Contribution — how much of your code is AI-generated
• Your Type — categorize them based on patterns:
  • "The Architect" — Write-heavy, designs from scratch
  • "The Scholar" — Read-heavy, studies code deeply
  • "The Hacker" — Shell-heavy, command-line warrior
  • "The Surgeon" — StrReplace-heavy, precise edits
  • "The Detective" — Grep-heavy, finds the clues
  • "The Explorer" — Glob-heavy, navigates the codebase
  • "The Orchestrator" — Task-heavy, delegates to subagents
  • "The Researcher" — SemanticSearch-heavy, finds by meaning

Keep it concise, punchy, fun. Something they'd screenshot and share.