用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/hummingbot/condor --skill xrpl-mm-deploy命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
基于 SOC 职业分类
| name | xrpl_mm_deploy |
| description | End-to-end playbook for standing up an XRPL CLOB maker — preflight, planner, controller deploy, verify. |
| when_to_use | When asked to set up, deploy, or launch market making on an XRPL pair. Follow start to finish when running as a delegate task. |
| created | 2026-07-28T00:00:00Z |
| source | agent:xrpl_market_maker |
Default path: controller mode (pmm_simple). Fall back to executor mode only after a
real controller failure (schema reject, deploy/status error, or no on-ledger orders).
explore_geckoterminal(action="top_pools", network="xrpl"). Need
real depth and turnover. RLUSD/XRP is currently the only pair with both.get_portfolio_overview(). Need free XRP for reserves (1 + 0.2×offers),
a trustline for the issued asset, and inventory on both sides.manage_routines(action="run", name="xrpl_mm_quote_planner",
strategy_id="xrpl_market_maker.rlusd_xrp_maker",
config={"xrpl_pair": "<pair>", "tick_interval_sec": <frequency_sec>,
"requote_interval_sec": <executor_refresh_time>,
"levels_per_side": <n>, "total_amount_quote": <capital>})
requote_interval_sec is the bot's executor_refresh_time in controller mode — not
frequency_sec. Wrong interval → false viable: false.
If viable: false with the right interval → do not deploy. Shorten
executor_refresh_time or wait. Never tighten below the floor. Large
divergence_vs_reference_bps usually means stale data, not opportunity.
pmm_simple only. Set leverage=1; leave stop_loss/take_profit/time_limit/
trailing_stop null.
Override these four defaults:
| Field | Default | Set to |
|---|---|---|
executor_refresh_time | 300 | 30 (from config) |
skip_rebalance | false | true |
buy_spreads / sell_spreads | 0.01,0.02 | planner controller_spreads (fractions) |
total_amount_quote | 100 | planner controller_total_amount_quote (XRP on RLUSD-XRP) |
pmm_simple centres on XRPL mid, not the CEX reference — if divergence is routinely
wide, recommend executor mode instead of deploying.
Two stores — update both when retuning later:
manage_controllers(action="upsert", target="config", ...)
manage_bots(action="deploy", ...)
Deploy under config bot_name (rlusd-xrp-maker) — stable name so restarts reattach.
Set max_global_drawdown_quote in the quote asset on every deploy.
Treat as failed → fall through to executors only when:
Journal the failure as category="execution". Falling back means clear bot_name to ''
so PnL attributes to direct executors.
manage_executors(action="create", controller_id="{agent_id}",
executor_config={"type": "order_executor", "connector_name": "xrpl", ...})
One LIMIT/LIMIT_MAKER level per side first; confirm on-ledger before laddering.
get_market_data(data_type="order_book", connector_name="xrpl", trading_pair="<pair>")
Offer must be visible at the expected price.
tecUNFUNDED_OFFER → sized against reserved XRP; recompute free balancetecNO_LINE / tecPATH_DRY → trustline missingState which is active: unhedged (deliberate XRP exposure) or hedged
(bitget_perpetual delta neutral — preferred).
dry_run → run_once → loop with short max_ticks and conservative limits.
execution_mode accepts exactly dry_run, run_once, or loop — nothing else.
There is no live value; continuous live trading is loop.