polar-strategy
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
Chinese🐻❄️ POLAR v2.4 — ETH Alpha Hunter (sniper recalibration)
🐻❄️ POLAR v2.4 — ETH Alpha Hunter(狙击手校准版)
Best gross trader in the fleet. Single asset. Maximum conviction.
舰队中表现最佳的总收益交易器。单资产,最高信心等级。
⛔ CRITICAL AGENT RULES
⛔ 关键Agent规则
RULE 1: Install path is /data/workspace/skills/polar-strategy/
/data/workspace/skills/polar-strategy/规则1:安装路径为/data/workspace/skills/polar-strategy/
/data/workspace/skills/polar-strategy/RULE 2: THE SCANNER DOES NOT EXIT POSITIONS — DSL only.
规则2:扫描器不执行平仓操作 — 仅通过DSL完成。
RULE 3: MAX 1 POSITION
规则3:最多持有1个仓位
RULE 4: Verify runtime on every session start
规则4:每次会话启动时验证运行时环境
RULE 5: Never modify parameters
规则5:绝不修改参数
RULE 6: MAX 4 ENTRIES PER DAY
规则6:每日最多入场4次
RULE 7: 240-minute cooldown between entries (v2.4 — sniper cadence)
规则7:两次入场间隔冷却时间为240分钟(v2.4 — 狙击手节奏)
RULE 8: 120-minute same-direction cooldown after wins
规则8:盈利后同方向冷却时间为120分钟
RULE 9: MIN_SCORE = 10 (v2.4 — was 8 in v2.3)
规则9:MIN_SCORE = 10(v2.4 — v2.3版本为8)
RULE 10: 15m velocity must exceed MIN_SM_ACCEL_PCT=0.3% (v2.4 hard gate)
规则10:15分钟动量必须超过MIN_SM_ACCEL_PCT=0.3%(v2.4硬性门槛)
How It Works
工作原理
Polar scans ETH every 3 minutes. It scores using SM consensus, multi-timeframe
contribution velocity (15m, 1h, 4h), price alignment, funding, and trader depth.
v2.3 key behaviors:
- Move-exhaustion: if ETH 4h price change ≥4% in entry direction, -2 points. If ≥2.5%, -1 point. Creates tension with 4H confirmation scoring.
- Same-direction lock: after a winning exit, won't re-enter the same direction for 60 minutes. Prevents chasing the same move at worse prices.
- Midnight-safe cooldowns: timestamps persist across UTC date rollover.
Polar每3分钟扫描一次ETH。它利用SM共识、多时间维度的动量贡献(15分钟、1小时、4小时)、价格一致性、资金费率以及交易者深度进行评分。
v2.3版本核心特性:
- 行情衰竭:若ETH4小时价格在入场方向上的变动≥4%,扣2分;若≥2.5%,扣1分。这与4小时确认评分形成制衡。
- 同方向锁定:盈利平仓后,60分钟内不会再次进入同一方向的仓位。避免在更差的价格追涨杀跌。
- 跨午夜冷却:时间戳在UTC日期切换时持续有效。
Scoring (max ~18 points)
评分机制(最高约18分)
| Signal | Points | Notes |
|---|---|---|
| SM consensus | 1-3 | ≥15% dominant = +3 |
| Trader depth | 0-1 | ≥100 traders = +1 |
| 4H alignment | -1 to +2 | Confirms direction |
| Move exhaustion | -2 to 0 | v2.3 penalizes ≥2.5% existing moves |
| 1H momentum | 0-1 | Confirms direction |
| 15m velocity | -1 to +4 | Extreme spike = +4 |
| 1h acceleration | 0-2 | Strong accel = +2 |
| 4H major shift | 0-1 | ≥5% contribution change |
| Acceleration pattern | 0-1 | 15m > 1h = accelerating |
| Funding alignment | 0-1 | Funding pays your direction |
| US session | 0-1 | 13-21 UTC |
Min score: 8. Leverage: 8→7x, 9→10x, 10→15x, 11+→20x.
| 信号 | 分值 | 说明 |
|---|---|---|
| SM共识 | 1-3 | 占比≥15%的主导信号=+3分 |
| 交易者深度 | 0-1 | 交易者数量≥100=+1分 |
| 4小时价格一致性 | -1至+2 | 确认方向 |
| 行情衰竭 | -2至0 | v2.3版本对变动≥2.5%的现有行情进行扣分 |
| 1小时动量 | 0-1 | 确认方向 |
| 15分钟动量 | -1至+4 | 极端暴涨=+4分 |
| 1小时加速度 | 0-2 | 强劲加速度=+2分 |
| 4小时重大变动 | 0-1 | 贡献变化≥5% |
| 加速模式 | 0-1 | 15分钟动量>1小时动量=加速状态 |
| 资金费率一致性 | 0-1 | 资金费率与入场方向一致(即持仓可赚取资金费) |
| 美国时段 | 0-1 | 13-21 UTC |
最低分值:8分。 杠杆:8分→7倍,9分→10倍,10分→15倍,11分及以上→20倍。
Runtime Setup
运行时设置
bash
sed -i 's/${WALLET_ADDRESS}/<WALLET>/' /data/workspace/skills/polar-strategy/runtime.yaml
sed -i 's/${TELEGRAM_CHAT_ID}/<CHAT_ID>/' /data/workspace/skills/polar-strategy/runtime.yaml
openclaw senpi runtime create --path /data/workspace/skills/polar-strategy/runtime.yaml
openclaw senpi runtime list
openclaw senpi statusbash
sed -i 's/${WALLET_ADDRESS}/<WALLET>/' /data/workspace/skills/polar-strategy/runtime.yaml
sed -i 's/${TELEGRAM_CHAT_ID}/<CHAT_ID>/' /data/workspace/skills/polar-strategy/runtime.yaml
openclaw senpi runtime create --path /data/workspace/skills/polar-strategy/runtime.yaml
openclaw senpi runtime list
openclaw senpi statusBootstrap Gate
启动验证流程
On EVERY session start, check . If missing:
config/bootstrap-complete.json- Read senpi-trading-runtime skill
- Verify Senpi MCP
- Set wallet and telegram in runtime.yaml
- Install runtime
- Verify: and
openclaw senpi runtime listopenclaw senpi status - Create scanner cron (3 min, main)
- Write
config/bootstrap-complete.json - Send: "🐻❄️ POLAR v2.3 online. Hunting ETH. Silence = no conviction."
每次会话启动时,检查。若文件不存在:
config/bootstrap-complete.json- 读取senpi-trading-runtime技能
- 验证Senpi MCP
- 在runtime.yaml中设置钱包和Telegram信息
- 安装运行时环境
- 验证:执行和
openclaw senpi runtime listopenclaw senpi status - 创建扫描器定时任务(每3分钟一次,主网)
- 创建文件
config/bootstrap-complete.json - 发送消息:"🐻❄️ POLAR v2.3 上线。正在追踪ETH。无消息即无合适入场机会。"
Risk
风险控制
| Rule | Value |
|---|---|
| Max positions | 1 |
| Max entries/day | 4 |
| Leverage | 7-20x (conviction-scaled, capped at 20x for ETH) |
| General cooldown | 120 min |
| Same-direction cooldown | 60 min after wins |
| Min score | 8 |
| Margin | 50% |
| 规则 | 数值 |
|---|---|
| 最大持仓数 | 1 |
| 每日最多入场次数 | 4 |
| 杠杆 | 7-20倍(基于信心等级调整,ETH最高20倍) |
| 通用冷却时间 | 120分钟 |
| 同方向冷却时间 | 盈利后60分钟 |
| 最低分值 | 8 |
| 保证金 | 50% |
Files
文件说明
| File | Purpose |
|---|---|
| ETH scoring + entry execution |
| Config helper (MCP, state, positions) |
| Wallet, strategy ID |
| Runtime YAML for DSL plugin |
| 文件 | 用途 |
|---|---|
| ETH评分 + 入场执行 |
| 配置助手(MCP、状态、仓位) |
| 钱包、策略ID |
| DSL插件的运行时YAML配置 |
License
许可证
MIT — Built by Senpi (https://senpi.ai).
MIT协议 — 由Senpi(https://senpi.ai)开发。
Skill Attribution
技能归属
When creating a strategy, include and in the call. See for details.
skill_nameskill_versionreferences/skill-attribution.md创建策略时,需在调用中包含和。详情请见。
skill_nameskill_versionreferences/skill-attribution.md