What is review-game-log?

The Review Game Log skill automates the extraction and formatting of game simulation data from local SQLite databases. It enables users to instantly retrieve comprehensive game summaries and turn-by-turn event logs using unique game IDs, transforming raw database records into structured, human-readable reports for deep tactical analysis and post-game reviews.

When should I use review-game-log?

review-game-log is useful in the following scenarios: • Post-Game Analysis: Quickly retrieve and summarize specific game sessions using UUIDs to evaluate match outcomes and player performance. • Tactical Auditing: Conduct turn-by-turn reviews to identify key turning points, specific card interactions, and player decision patterns. • Simulation Debugging: Filter and analyze specific game phases or event types to identify logic errors or unexpected behaviors in game simulations. • Data Reporting: Generate structured tables of saved game data, including player names, final results, and total log entries for archival or sharing purposes.

name	review-game-log
description	Automatically invoked when reviewing game simulation logs. Retrieves saved game data by ID and formats it for analysis.

Review Game Log Skill

This skill activates when you need to review a saved MTG game simulation log.

When This Skill Activates

Automatically engage when:

User asks to review a game log by ID
User wants to analyze a saved game
User provides a game ID (UUID format) and asks about game events

Workflow

1. Identify Game ID

The game ID is a UUID like a1b2c3d4-e5f6-7890-abcd-ef1234567890. Users may provide:

Full UUID
Partial UUID (first 8 characters)
Reference like "the game I just saved"

2. Fetch Game Data

Use the Bash tool to invoke the Tauri command via the app's API or read directly from the SQLite database:

# Option 1: Query SQLite directly (preferred for Claude)
sqlite3 ~/.local/share/com.scrye3.app/scrye3.db "
SELECT
  s.game_id,
  s.player1_name,
  s.player2_name,
  s.winner,
  s.final_turn,
  s.total_log_entries,
  s.saved_at
FROM saved_games s
WHERE s.game_id LIKE 'GAME_ID_PREFIX%'
LIMIT 1;
"

# Then fetch logs
sqlite3 ~/.local/share/com.scrye3.app/scrye3.db "
SELECT
  timestamp,
  event_type,
  turn_number,
  phase,
  step,
  message,
  details
FROM game_logs
WHERE game_id = 'FULL_GAME_ID'
ORDER BY timestamp ASC, id ASC;
"

3. Format Output

Present the game summary and key events:

## Game Summary
- **Game ID**: {game_id}
- **Players**: {player1_name} vs {player2_name}
- **Result**: {winner} won on turn {final_turn}
- **Total Events**: {total_log_entries}
- **Saved**: {saved_at}

## Key Events

| Turn | Phase | Event | Details |
|------|-------|-------|---------|
| 1 | beginning | Game started | ... |
| 1 | precombat_main | Land played | Forest |
| ... | ... | ... | ... |

4. Analysis

After presenting the log, offer to:

Summarize the game flow
Identify key turning points
Analyze player decisions
Find specific events (combat, spells cast, etc.)

Constraints

Only access games that have been explicitly saved
Game logs are read-only (cannot modify saved games)
Database path may vary by platform:
- Linux: ~/.local/share/com.scrye3.app/scrye3.db
- macOS: ~/Library/Application Support/com.scrye3.app/scrye3.db
- Windows: %APPDATA%/com.scrye3.app/scrye3.db

Examples

User: "Review game a1b2c3d4"

Claude:

Query database for game starting with a1b2c3d4
Present summary and log table
Offer analysis options

User: "What happened in turn 5 of that game?"