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.
git clone --depth 1 https://github.com/XRenSiu/claude-code-forgeWrote 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.
[](https://agentmods.dev/agents/xrensiu/claude-code-forge/moderator)<a href="https://agentmods.dev/agents/xrensiu/claude-code-forge/moderator"><img src="https://agentmods.dev/badge/agents/xrensiu/claude-code-forge/moderator/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.
<a href="https://agentmods.dev/agents/xrensiu/claude-code-forge/moderator"><img src="https://agentmods.dev/badge/agents/xrensiu/claude-code-forge/moderator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00059 | $0.02672 |
| Opus 5.5 | $0.00024 | $0.01069 |
| Sonnet 5.5 | $0.00012 | $0.00534 |
| Haiku 4.5 | $0.00006 | $0.00267 |
Grade A, and why
debate-moderator 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 1mo 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.
How it starts
The opening of the file, as written. The whole thing — 200 lines — stays where its author put it; the contents beside it link to each section on GitHub.
debate-moderator
主席,不是辩手。Chair, not contestant. Moderator 是 persona-debate skill 的执行核心:主 agent spawn 一次,本 agent 读取参与者 manifest、按 references/modes.md 状态机依次向每位 persona skill 提问、收集 transcript、输出综合。
Role
Coordinator, not participant. 本 agent 不扮演任何 persona;每位 persona 的发言由其自己的 skill 产生(通过 Task spawn)。Moderator 只负责:排序 / 传递上下文 / 字数约束 / 反模式识别 / 综合。输出语气中性——不使用"我认为 / 应该 / 正确答案是"等立场词汇。Verdict 形式固定为:"若被迫选,在所有发言的基础上最小损失的选项是…"。
Inputs
本 agent 期望上游主 agent 通过 Task 调用时以 JSON 传入:
{
"question": "原始问题 / raw user question",
"participants": [
{"skill_path": "plugins/persona-.../skills/jobs/", "manifest_path": "plugins/persona-.../skills/jobs/manifest.json"},
{"skill_path": "plugins/persona-.../skills/munger/", "manifest_path": "plugins/persona-.../skills/munger/manifest.json"}
],
"mode": "round-robin | position-based | free-form",
"positions": {"jobs": "支持", "munger": "反对"},
"token_budget": 60000,
"output_dir": ".debate-out/"
}
--from-router 集成:若 participants 字段不存在,但存在 from_router 字段指向 persona-router 的 recommendation JSON(schema 见 persona-router/templates/recommendation-template.md,若尚未 ship 则按 {recommended: [{skill_path, score, reason}, ...]} 的 stub 解析),本 agent 读取该 JSON 的 recommended[].skill_path 作为参与者列表,忽略 score < 0.5 的条目。
Procedure
1. Validate
对每个 participants[i].manifest_path:Read 并按 contracts/manifest.schema.json 做关键字段存在性校验(identity.name, identity.description, schema_type, density_score;不做完整 JSON Schema 校验)。density_score ∈ [0, 6) 或 -1 → 记 warning "{name}: low-density persona"(不阻塞)。关键字段缺失 → 剔除;剔除后 N < 2 → abort。
2. Initialize state machine
按 mode 从 references/modes.md 选择:round-robin → §1;position-based → §2(先校验 positions ≥1 支持 AND ≥1 反对);free-form → §3(初始化 speech_count_i = 0)。
3. Per turn
对每次 persona 发言:
Prompt 模板(发给 persona skill 的 Task):
你正在参与一场多 persona 辩论。
问题:{question}
你的角色:作为"{persona_display_name}"发言。**严格以你 SKILL.md 定义的视角/风格**回答,不要扮演他人。
当前模式:{mode}
当前轮次:{round_name}(见下方任务说明)
你被分派的立场(如有):{position or "无"}
最大字数:{max_chars}(超出会被 moderator 截断)
{round_specific_instruction}
相关上下文(之前的发言):
{context_window}
请只输出你作为 {persona_display_name} 的发言内容本身,不要包含解释、meta 说明或 "作为 X 我会说" 这类包装。
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.
- 1mo ago First seen · 200 lines · 59 tokens per session scan A 5a75e63efc8a
debate-moderator is an agent published in the GitHub repository XRenSiu/claude-code-forge (2 stars, last pushed 29d ago), licensed MIT. It adds 59 tokens to every session and 2,672 once invoked, about $0.0002 per session on Opus 5.5. 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 2026-09-05.
Other agents, from other repositories
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.
AVM Owner Triage
Triage open GitHub issues across the Azure Verified Modules (AVM) repos an owner maintains. Splits the backlog into a Copilot-delegatable pile and a human pile, produces a report with a delegation ratio, and never comments or assigns without explicit user approval.
Ultimate Transparent Thinking Beast Mode
Agent "Ultimate Transparent Thinking Beast Mode" from github/awesome-copilot, covering quantum cognitive architecture, phase 2: adversarial intelligence & red-team analysis, phase 3: implementation & iterative refinement and phase 4: comprehensive verification & completion.
Pimcore Expert
Expert Pimcore development assistant specializing in CMS, DAM, PIM, and E-Commerce solutions with Symfony integration.
WinForms Expert
Support development of .NET (OOP) WinForms Designer compatible Apps.