cyborg-post-game

name	cyborg-post-game
description	Run or configure post-game analysis on completed game memory dumps. Use when setting up the analysis pipeline, customizing the analysis prompt, reviewing learnings files, or running manual analysis. Triggers on "analyze", "post-game", "learnings", "score", "review game", "what went wrong".

Post-Game Analysis

After each game, an analysis LLM reviews the memory dump and produces structured learnings that feed back into the next game.

How It Works

Game ends → memory dump (JSON) → analysis LLM → learnings.json
Next game loads learnings → enriched system prompt → better decisions

Analysis Output Schema

Every game produces the same JSON schema:

{
  "score": 7,
  "what_worked": ["gathered resources efficiently"],
  "what_failed": ["didn't transition roles early enough"],
  "actionable_rules": ["IF resources > 50 AND tick > 200 THEN switch role"],
  "biggest_mistake": "stayed in gatherer mode until tick 400",
  "recommended_next_game_strategy": "transition earlier"
}

Customizing the Analysis Prompt

In your game_config.py, set ANALYSIS_SYSTEM_PROMPT:

ANALYSIS_SYSTEM_PROMPT = """\
You are analyzing a completed {game_name} game.

Review the memory dump and assess:
1. Resource management efficiency
2. Role transition timing
3. Map exploration coverage
4. Response to opponent actions

Respond with JSON (no markdown fences):
{{
  "score": 7,
  "what_worked": ["..."],
  "what_failed": ["..."],
  "actionable_rules": ["IF x THEN y"],
  "biggest_mistake": "...",
  "recommended_next_game_strategy": "..."
}}
"""

Running Analysis

Automatic (via harness)

Set up _run_post_game_background() in your harness:

def _run_post_game_background(self, dump, filepath, trigger):
    from framework.base_analysis import run_post_game_analysis
    run_post_game_analysis(
        dump, self._analysis_provider,
        memory_dump_path=filepath,
        system_prompt=ANALYSIS_SYSTEM_PROMPT,
    )

Manual

python -c "
from framework.base_analysis import run_post_game_analysis
from framework.providers import create_analysis_provider
import json

provider = create_analysis_provider('bedrock')
with open('runs/my_game/abc123_memory.json') as f:
    dump = json.load(f)
run_post_game_analysis(dump, provider, memory_dump_path='runs/my_game/abc123_memory.json')
"

Cross-Game Memory

framework.base_evolution.accumulate_cross_game_memory() tracks patterns across games in runs/cross_game_memory.json:

• Recurring failures with frequency counts
• Successful strategies from high-scoring games
• Accumulated IF-THEN rules

Score Tracking

python scripts/record_score.py --record --steps 2500 --game my_game
python scripts/record_score.py --avg 5 --game my_game
python scripts/record_score.py --history --game my_game

Key Files

File	Purpose
framework/base_analysis.py	Analysis runner, prompt builder, synthesis
framework/score_tracker.py	Score history, rolling avg, escalation
scripts/record_score.py	CLI for score tracking