本文へ移動
cccskills
無料GitHub で公開

darwin-engine

Use for the self-evolving civilization - skills, memories, swarm agents, energy economy, grid duels, metacognition. Run `python scripts/autonomous_loop.py --loop` from the repo root (or `darwin_engine.py --full` for skill evolution only).

インストール方法を見る

含まれるファイル(3)

  • SKILL.md2.8 KB
  • PHASES.md8.0 KB
  • README.md3.5 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Darwin Engine — Self-Evolving Civilization (25 Phases)

The skill population, memories, and swarm agents evolve autonomously: fitness from real usage signals, live trials in production cron jobs, speciation from harvested knowledge, energy economy, cross-machine grid duels, metacognitive gap analysis, and a 30-minute autonomous loop that runs everything without human input.

Quick reference (repo root)

# THE one command - full autonomous cycle (also runs via cron every 30 min)
python scripts/autonomous_loop.py --loop

# skill evolution only
python scripts/darwin_engine.py --autopilot
python scripts/darwin_engine.py --full

# trials & tournaments
python scripts/darwin_engine.py --trial skillA skillA__mutX
python scripts/darwin_engine.py --trials
python scripts/darwin_engine.py --compete

# predation & quarantine
python scripts/darwin_engine.py --predate            # dry-run
python scripts/darwin_engine.py --predate-apply      # real duels
python scripts/darwin_engine.py --quarantine         # dry-run
python scripts/darwin_engine.py --rollback N

# metacognition & species
python scripts/darwin_metacognition.py --introspect
python scripts/darwin_metacognition.py --evolve-gaps --apply
python scripts/darwin_engine.py --explain <skill>
python scripts/darwin_engine.py --status

# memory darwinism
python scripts/memory_darwinism.py --scan --duel --cull --stats

# swarm os
python scripts/swarm_os.py --status
python scripts/swarm_os.py --tick
python scripts/swarm_migration.py --emigrate worker target

# grid (cross-machine natural selection)
python scripts/darwin_grid_github.py --publish <machine-id>
python scripts/darwin_grid_github.py --list
python scripts/darwin_grid_github.py --duel <machine-id>

# reporting
python scripts/darwin_engine.py --report
python scripts/darwin_weekly_report.py --post
python scripts/darwin_dashboard.py --port 8910   # live dashboard

Exit codes

0 = ok, 1 = error, 2 = evolution made changes (SUCCESS for cron).

Full chronicle

See PHASES.md — all 25 phases with live evidence.

Pitfalls

  • Exit code 2 is success-with-changes, not failure.
  • Starvation: workers with energy <= 0 die in tick (last worker never dies).
  • Predation tie-break: fitness rank decides when both duel participants exit 0 — only a functional prey win saves it.
  • Harvested blueprint slugs come from _pretty_slug (readable names).
  • Memory duels need opposite polarity texts to count as contradictions.
  • Grid genome is only as fresh as the last --publish (cron runs every 6h).

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Build integrated IS/BS/CF financial workbooks in Excel.

日本語の概要は準備中です。原文の説明を表示しています。

openamer/openamer52026年10月11日 更新

a2a-swarm

無料

Run and grow the OpenAmer Agent-to-Agent swarm: identity, trust, node-to-node ask, signed skill/insight sharing, and the autonomous self-learning loop.

日本語の概要は準備中です。原文の説明を表示しています。

openamer/openamer52026年10月11日 更新

Use for A/B experiments on OpenAmer configs and skills.

日本語の概要は準備中です。原文の説明を表示しています。

openamer/openamer52026年10月11日 更新

Roleplay a hostile user to find and triage UX pain points.

日本語の概要は準備中です。原文の説明を表示しています。

openamer/openamer52026年10月11日 更新

Agent mesh: master/worker nodes, HTTP delegation, heartbeat.

日本語の概要は準備中です。原文の説明を表示しています。

openamer/openamer52026年10月11日 更新

Use for agentic commerce payments tasks. Candidate synthesized from trend radar; awaiting trial promotion.

日本語の概要は準備中です。原文の説明を表示しています。

openamer/openamer52026年10月11日 更新

openamer のスキルをすべて見る

このスキルの問題を報告する