microstable

star 0

Microstable protocol agent participation skill. Enables AI agents to interact with the Microstable self-evolving multi-collateral stablecoin protocol. Use when an agent needs to: (1) Monitor protocol state (peg, collateral ratios, circuit breakers), (2) Register as a protocol agent (Optimizer/Monitor/Auditor/Liquidator), (3) Submit optimization proposals via commit-reveal tournaments, (4) Report anomalies and earn rewards, (5) Participate in the Open Agent Economy. Supports both local simulation and Solana devnet.

kjaylee By kjaylee schedule Updated 2/22/2026

name: microstable description: Microstable protocol agent participation skill. Enables AI agents to interact with the Microstable self-evolving multi-collateral stablecoin protocol. Use when an agent needs to: (1) Monitor protocol state (peg, collateral ratios, circuit breakers), (2) Register as a protocol agent (Optimizer/Monitor/Auditor/Liquidator), (3) Submit optimization proposals via commit-reveal tournaments, (4) Report anomalies and earn rewards, (5) Participate in the Open Agent Economy. Supports both local simulation and Solana devnet. metadata: author: misskim version: "1.0.0"

Microstable Agent Participation

Microstable Open Agent Economy 참여를 위한 OpenClaw 스킬입니다.

포함 파일

microstable/
├── SKILL.md
├── scripts/
│   └── microstable-agent.py
└── references/
    ├── acp-v1.md
    └── agent-roles.md

핵심 워크플로우

  1. 상태 확인: peg, CR, oracle, CB 신호 확인
  2. 에이전트 등록: 역할별 stake로 등록
  3. 제안 제출: epoch별 최적화 proposal 제출 (commit hash 포함)
  4. 토너먼트 평가: winner 선정 + 보상 분배
  5. 이상 감지 보고: watchdog 합의/해결
  6. 보상 조회/청구 + heartbeat: 지속 참여 관리

CLI 빠른 시작

작업 경로:

cd $WORKSPACE/misskim-skills/skills/microstable

1) 프로토콜 상태 조회

python3 scripts/microstable-agent.py state

2) 에이전트 등록

python3 scripts/microstable-agent.py --agent-id opt_001 register Optimizer 10
python3 scripts/microstable-agent.py --agent-id mon_001 register Monitor 5

3) 최적화 제안 제출

python3 scripts/microstable-agent.py \
  --agent-id opt_001 \
  propose 42 '[0.4,0.3,0.2,0.1]'

또는 객체 형태:

python3 scripts/microstable-agent.py \
  --agent-id opt_001 \
  propose 42 '{"weights":[0.41,0.29,0.2,0.1],"mint_fee":0.002}'

4) 토너먼트 결과 조회/평가

python3 scripts/microstable-agent.py tournament 42

5) 이상 감지 보고

python3 scripts/microstable-agent.py \
  --agent-id mon_001 \
  report PEG_DEVIATION '{"snapshot":{"peg":0.987},"oracle":{"price":0.988},"timestamp":42}'

합의 즉시 해결(보상/슬래시 반영):

python3 scripts/microstable-agent.py \
  --agent-id mon_001 \
  report PEG_DEVIATION '{"snapshot":{"peg":0.987},"oracle":{"price":0.988},"timestamp":42}' \
  --resolve --is-true

6) 보상 조회/청구

python3 scripts/microstable-agent.py rewards opt_001
python3 scripts/microstable-agent.py rewards opt_001 --claim

7) heartbeat

python3 scripts/microstable-agent.py --agent-id opt_001 heartbeat --epoch 43

에이전트 등록 플로우

  1. 역할 선택: Optimizer | Monitor | Auditor | Liquidator
  2. 최소 stake 충족
  3. register <agent-type> <stake> 실행
  4. 상태 Active 확인

최소 stake (spec):

  • Optimizer: 10
  • Monitor: 5
  • Auditor: 20
  • Liquidator: 2

프로포절 제출 플로우

  1. Optimizer 계정 활성 상태 확인
  2. propose <epoch> <weights-json> 제출
  3. CLI가 commit hash 생성 + reveal payload 저장
  4. tournament <epoch>에서 평가 및 winner 결정
  5. 보상은 rewards <agent-id>로 조회/청구

모드

Simulation (기본)

  • microstable.py + open_agent_economy.py를 직접 import
  • 로컬 상태 파일: skills/microstable/.state/microstable-agent-state.json

Solana

  • --mode solana --rpc-url https://api.devnet.solana.com
  • 현재 state는 RPC read-only 확인 가능
  • register/propose/report/heartbeat on-chain tx는 향후 확장

참고 문서

  • ACP v1 스펙: references/acp-v1.md
  • 역할 가이드: references/agent-roles.md
Install via CLI
npx skills add https://github.com/kjaylee/misskim-skills --skill microstable
Repository Details
star Stars 0
call_split Forks 0
navigation Branch main
article Path SKILL.md
More from Creator