What is llm-arbitration-policy?

A rule set for an orchestrator that combines recommendations from several specialist agents into one stock-trading decision. It also checks whether those agents failed, timed out, or produced unusable results.

Use it to combine bullish, bearish, and neutral signals, account for confidence and importance, record data gaps, and produce a decision in a required JSON format.

It gives conflicting or incomplete recommendations a consistent review process and can choose to wait or hold when the evidence is weak. It does not fetch the original market data or replace specialist analysis.

llm-arbitration-policy — Skill for Claude Code and Codex by duolongwor

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add duolongworld/AI_Renaissance --skill llm_arbitration_policy

Clone the repo

git clone --depth 1 https://github.com/duolongworld/AI_Renaissance

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/duolongworld/ai_renaissance/llm_arbitration_policy/github.svg)](https://agentmods.dev/skills/duolongworld/ai_renaissance/llm_arbitration_policy)

Your own site

<a href="https://agentmods.dev/skills/duolongworld/ai_renaissance/llm_arbitration_policy"><img src="https://agentmods.dev/badge/skills/duolongworld/ai_renaissance/llm_arbitration_policy/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/duolongworld/ai_renaissance/llm_arbitration_policy"><img src="https://agentmods.dev/badge/skills/duolongworld/ai_renaissance/llm_arbitration_policy.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 42 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,908 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits

NVIDIA SkillSpector pass 7 Sept 2026

How audits are shown

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00042	$0.01908
Opus 5	$0.00021	$0.00954
Sonnet 5	$0.00008	$0.00382
Haiku 4.5	$0.00004	$0.00191

Measured 12d ago against content hash 3c3503f4951b, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

llm-arbitration-policy scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/orchestrator/llm_arbitration_policy/SKILL.md · 164 lines

How it starts

The opening of the file, as written. The whole thing — 164 lines — stays where its author put it; the contents beside it link to each section on GitHub.

LLM 仲裁策略 Skill

1. 适用范围

所属小组：开发2组（Orchestrator）

适用任务：

汇总多个专家 Agent 产出的 SignalBundle
结合编排 trace 检查专家执行失败、超时、无效信号等上下文
在必要时调用 MCP 工具补充市场、风险或外部上下文
输出兼容 agents.orchestrator.arbitration.ArbitrationResult 的最终仲裁结果

边界说明：

本 Skill 只定义 LLM 仲裁方法和输出约束，不直接获取原始行情、财报、新闻或股吧数据
不替代专家 Skill 的专业判断；专家 Signal 是主要输入，外部 MCP 信息只作为补充证据
当专家信号数量不足、方向严重冲突或 MCP 工具不可用时，应降低 confidence，必要时输出 wait 或 hold
不得输出超出 schema 的交易指令，例如具体买入价格、止损价或卖出批次

2. 输入材料

必填输入

stock_code：股票代码
signal_bundle：专家 Agent 产出的信号集合，包含方向、置信度、权重、推理和来源
execution_trace：Orchestrator 编排 trace，包含专家执行状态、失败数、超时数、无效数等
available_skills：当前框架加载到的 Skill 列表
required_output_schema：必须遵守的最终输出结构

可选输入

mcp_tools：可用 MCP 工具列表及注册信息
MCP 工具返回的市场上下文、风险事件、数据质量说明或其他外部证据

缺失处理

如果 signal_bundle.signals 为空，输出 decision: "wait"、direction: "neutral"，confidence 不高于 0.3，position_ratio 为 0
如果多数专家执行失败或超时，在 risks 中说明数据覆盖不足，并降低 confidence
如果 MCP 工具不可用，不应阻塞仲裁，但必须在 risks 或 reasoning_chain 中说明外部验证不足

3. 仲裁流程

按下面步骤执行：

读取 signal_bundle，按方向统计 bullish、bearish、neutral 信号数量。
结合每个信号的 confidence 与 weight，评估看多、看空和中性证据强度。
检查 execution_trace，识别专家失败、超时、无效输出和数据覆盖缺口。
如存在可用 MCP 工具，优先补充与当前标的相关的市场上下文和风险事件。
判断主方向：证据明显偏多时为 bullish，明显偏空时为 bearish，证据冲突或不足时为 neutral。
计算 confidence：方向一致性越高、专家覆盖越完整、证据质量越高，置信度越高；反之降低。
计算 position_ratio：只在 direction 为 bullish 或 bearish 且 confidence 足够时给出非零仓位；中性或等待时为 0。
输出最终 JSON，不添加 Markdown 解释或额外文本。

4. 判断规则

方向判断

多数高置信度加权信号为 bullish，且无重大风险抵消时，输出 direction: "bullish"
多数高置信度加权信号为 bearish，或风险类信号强烈负面时，输出 direction: "bearish"
多空信号接近、核心专家缺失、证据质量不足或外部上下文不明确时，输出 direction: "neutral"

决策映射

direction: "bullish" 且 confidence >= 0.6 时，通常输出 decision: "buy"
direction: "bearish" 且 confidence >= 0.6 时，通常输出 decision: "sell"
direction: "neutral" 或多空证据冲突时，通常输出 decision: "hold"
输入不足、执行失败较多、外部验证缺失或风险不可判断时，输出 decision: "wait"

仓位约束

position_ratio 必须在 0 到 1 之间
decision 为 hold 或 wait 时，position_ratio 应为 0
非零仓位应随 confidence、信号一致性和风险水平调整
当风险提示包含重大不确定性、数据缺失或专家执行异常时，应主动下调仓位

Read the full file on GitHub · 164 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

12d ago First seen · 164 lines · 42 tokens per session scan A 3c3503f4951b

Subscribe to this mod's changes

llm-arbitration-policy is a skill published in the GitHub repository duolongworld/AI_Renaissance (59 stars, last pushed 14d ago), licensed Apache-2.0. It adds 42 tokens to every session and 1,908 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Building agents Finance and fintech

sector-rotation

An analysis framework for comparing industries in the Chinese A-share stock market, using business conditions, price momentum, valuation, and money flows. It produces rankings and higher- or lower-allocation suggestions.

HKUDS/Vibe-Trading · 39 tokens

strategy-pivot-designer

Detect backtest iteration stagnation and generate structurally different strategy pivot proposals when parameter tuning reaches a local optimum.

tradermonty/claude-trading-skills · 28 tokens

twitter-reader

Read Twitter/X for financial research using opencli (read-only). Use this skill whenever the user wants to read their Twitter feed, search for financial tweets, view bookmarks, look up user profiles, or gather market sentiment from Twitter/X. Triggers include: "check my feed", "search Twitter for", "show my…

himself65/finance-skills · 161 tokens

llm-arbitration-policy