build-consensus
About
This skill enables distributed consensus without central authority using mechanisms like threshold voting and quorum sensing. It's designed for scenarios where groups need to make decisions without a designated leader or when centralized decision-making creates bottlenecks. Developers can apply it to integrate diverse stakeholder perspectives or build automated consensus systems like distributed databases and multi-agent AI.
Quick Install
Claude Code
Recommendednpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/build-consensusCopy and paste this command in Claude Code to install this skill
Documentation
建共識
無中央而於散將間達共—以偵倡、閾法定、委動—模於蜂群決。
用
- 群當擇諸選而無定導
- 中央決成瓶或單點敗
- 當事人有異訊與視角當整
- 昔決受群思(早收)或析癱(無收)
- 設自動系需達共(散庫、多將 AI)
- 輔
coordinate-swarm於協需顯群決時
入
- 必:當決之事(二擇、N 中擇、參設)
- 必:參將(隊員、服、投者)
- 可:知選附初質評
- 可:決急(時預)
- 可:可容誤率(群偶擇次佳乎?)
- 可:當決敗模(群思、困、振)
行
一:獨偵生議
顯倡前先充探決空。
- 派偵獨探選空:
- 每偵評選而不知他偵所得
- 獨評防早聚於流而庸之選
- 偵數:最少每嚴選三偵(以穩)
- 偵出結構評:
- 選 ID
- 質分(歸一 0-100 或類:劣/普/優/卓)
- 所識要強與險
- 信級(此選多徹評?)
- 合偵報而不濾—過最低質閾之諸選皆入倡階
得: 獨評諸議附質分與評。無選由單評除;視角多存。
敗: 偵聚同選而非獨評→偵非真獨。以顯訊障重行。太多選過倡階→升最低質閾。太少→降或加偵。
二:行倡動(搖擺舞)
令偵倡所愛選,倡強依質比。
- 每偵倡己頂選:
- 倡強比質分(佳選得更烈倡)
- 倡公—諸將觀諸倡號
- 倡呈證與質評,非但偏
- 未委將觀倡而評:
- 跟倡選而獨察
- 己察證質→加入倡
- 察揭質低於所宣→不加
- 互察動:
- 弱選倡自失從—將獨驗而棄
- 強選倡因證質而得從
- 程自修:誇倡敗於驗步
Advocacy Dynamics:
┌─────────────────────────────────────────────────────────┐
│ Scout A advocates Option 1 (quality 85) ──→ ◉◉◉◉◉ │
│ Scout B advocates Option 2 (quality 70) ──→ ◉◉◉ │
│ Scout C advocates Option 3 (quality 45) ──→ ◉ │
│ │
│ Uncommitted agents inspect: │
│ Agent D inspects Option 1 → confirms → joins ◉◉◉◉◉◉ │
│ Agent E inspects Option 2 → confirms → joins ◉◉◉◉ │
│ Agent F inspects Option 3 → disagrees → inspects Opt 1│
│ → confirms → joins ◉◉◉◉◉◉◉│
│ │
│ Over time: Option 1 advocacy grows, Option 3 fades │
└─────────────────────────────────────────────────────────┘
得: 佳選倡經時而長—將獨驗質。弱選倡因驗敗而衰。群自聚於強選而無將令擇。
敗: 倡不聚(兩選並駕)→選或真等—以閾進或用破平律。倡過速聚於庸選→增評獨(多偵、嚴訊障)加強互察步。
三:設法定閾而委
定觸集行之委閾。
- 設法定閾:
- 簡決:將之五成 + 1 委一選
- 重決:66-75% 委一選
- 危/不可逆:80%+ 委一選
- 大法:重→閾高→共緩而穩
- 監委積:
- 跟每選委將數經時
- 透示委級(諸將見當態)
- 週中勿允撤委(防振)
- 法定達:
- 勝選為集決而採
- 敗選倡者認決(無叛將)
- 施即始—共後延耗委
得: 清法定時—足將獨委一選。決正因生於獨評,非權或強。
敗: 時預內法定不達→升步四(困解)。法定達而將不悅→倡階過短—將未充評而委。共事後發現誤→獨偵不足—次週增偵異與評徹。
四:解困
自然共程滯→破決堵。
- 斷困類:
- 真平:兩選等佳→擲錢;延價逾擇「誤」等選
- 訊缺:將不能充評→增偵而後重行倡
- 派成:固派拒互察→加必轉令倡者察對選
- 選泛:過多選散委→除底五成重行倡
- 施宜解:
- 真平:隨擇或合選若容
- 訊缺:時限偵展
- 派成:迫互察輪
- 選泛:排名淘
- 解後重置法定鐘,重行步三
得: 困以宜預解。解顯而群認為公程,雖個將欲異果。
敗: 同決反復困→決框或誤。退問:決可分小獨決乎?範可減乎?有「試兩觀之」選乎?時共之佳乃「行時限試」。
五:評共質
評程生佳決,非但決。
- 決後評:
- 勝選由至少 N 將獨驗乎?
- 決速宜乎(不快/群思、不緩/癱)?
- 程揭單決者或遺之訊乎?
- 將委施乎,抑但從?
- 跟共健指:
- 達法定時:經續決而減示學;增示繁或失能
- 偵委比:每委需多偵?高比 = 繁決或低信
- 決後悔率:群多欲擇異?
- 學回程:
- 依決要與昔準調法定閾
- 依選繁調偵數
- 依昔達法定時調時預
得: 回饋環經時改共質。群學更效偵、更誠倡、更信委。
敗: 共質指劣(高悔、緩決)→審程之構敗:偵異不足、倡無驗、閾低於決類。重建具敗階,非修全程。
驗
- 議以獨偵生(無隨群)
- 倡強依所評質比
- 未委將獨驗倡選
- 法定閾宜於決要
- 法定達且決速施
- 困解機可用(雖未用)
- 決後質評已行
忌
- 略獨偵:直跳倡生群思。共質全依獨評之質
- 平倡不平選:諸選無論質皆同倡→程退為隨擇。倡必比所評質
- 撤委:允將撤委生振。週中委→諸將守至週解
- 混共與全同:共需足同,非全同。候 100%→永困
- 略敗側:倡敗選之將有群需之訊。其慮當導施,雖不阻決
參
coordinate-swarm— 支號基共機之基協框defend-colony— 集護決常需威下速共scale-colony— 群大顯變時共機當適dissolve-form— 形技控解,解前共至要plan-sprint— 衝規涉隊共於委範conduct-retrospective— 復盤乃程改共之一build-coherence— AI 自用變;映蜂群democracy 於單將多路推附信閾與困解
GitHub Repository
Related Skills
content-collections
MetaThis skill provides a production-tested setup for Content Collections, a TypeScript-first tool that transforms Markdown/MDX files into type-safe data collections with Zod validation. Use it when building blogs, documentation sites, or content-heavy Vite + React applications to ensure type safety and automatic content validation. It covers everything from Vite plugin configuration and MDX compilation to deployment optimization and schema validation.
polymarket
MetaThis skill enables developers to build applications with the Polymarket prediction markets platform, including API integration for trading and market data. It also provides real-time data streaming via WebSocket to monitor live trades and market activity. Use it for implementing trading strategies or creating tools that process live market updates.
creating-opencode-plugins
MetaThis skill helps developers create OpenCode plugins that hook into 25+ event types like commands, files, and LSP operations. It provides the plugin structure, event API specifications, and implementation patterns for JavaScript/TypeScript modules. Use it when you need to intercept, monitor, or extend the OpenCode AI assistant's lifecycle with custom event-driven logic.
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
