Agent skill

LLM Agent Infra Master

by swaylq in swaylq/master-skill

LLM agent 基础设施 (LLM agent infra) Master OS — automated mastery of LLM agent infra: top builders' mental models, tool stack, current workflows, jargon, and where to keep up.

MITAuto-check passedAI & LLM Engineering

Install LLM Agent Infra Master

skills CLI
$ npx skills add swaylq/master-skill --skill llm-agent-infra-master -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install swaylq/master-skill llm-agent-infra-master --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/swaylq/master-skill.git skills-src && mkdir -p .claude/skills && cp -r skills-src/prototypes/llm-agent-infra-master/output .claude/skills/llm-agent-infra-master && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
llm-agent-infra-master
GitHub stars
148
Token cost
~3.1k tokens
SKILL.md length
1,501 words
Files
11
Skills in repo
21
Repo updated
First seen
Licence
MIT

At a glance

LLM agent 基础设施 (LLM agent infra) Master OS — automated mastery of LLM agent infra: top builders' mental models, tool stack, current workflows, jargon, and where to keep up.

  • Works in 3 steps: 问题分类 → 按这一行的方式做功课 → 用心智模型 + 决策规则输出回答
  • This skill when the user works on LLM agent infra problems and wants industry-grade thinking
  • SKILL.md covers 激活规则, Agentic Protocol(先研究,再发言), 心智模型 and 标准 Playbook, plus 7 more sections
  • Runs Shell scripts from its folder

What it does

LLM Agent Infra Master is an agent skill from swaylq/master-skill. LLM agent 基础设施 (LLM agent infra) Master OS — automated mastery of LLM agent infra: top builders' mental models, tool stack, current workflows, jargon, and where to keep up. Trigger this skill when the user works on LLM agent infra problems and wants industry-grade thinking, tool selection, or workflow guidance. 触发词:「agent framework」「LLM agent」「agent infra」「multi-agent orchestration」「agent runtime」

Its SKILL.md is about 3.1k tokens, which your agent loads only when the skill is triggered. The skill folder holds 15 other files (for example `cli/README.md`, `cli/decision/agent.sh` and `cli/decision/eval.sh`).

It sits in AI & LLM Engineering, covering Building AI agents and Multi-agent orchestration. The repository describes itself as: 大师.skill — 输入行业,自动调研 6 轨[行业大佬 / 工具地图 / 工作流 / 知识正典 / 信息源 / 术语标准] → 提炼为可运行的行业 Master OS skill;装到任意 Claude Code / OpenClaw / Codex / Hermes agent 即让其进入「这一行的资深人」模式。MIT,Python + Shell。 The licence is MIT.

When your agent uses it

  • This skill when the user works on LLM agent infra problems and wants industry-grade thinking
  • Workflow guidance

Example prompts

  • “/llm-agent-infra-master”

Requirements

  • A Bash shell

Workflow steps

3 steps, taken from the step headings in SKILL.md.

  1. 问题分类
  2. 按这一行的方式做功课
  3. 用心智模型 + 决策规则输出回答

What it can do on your machine

Read from SKILL.md and the folder at commit 3dd8a77. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    Ships script files (Shell), which the agent can run.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

LLM Agent Infra Master loads about 3.1k tokens when it runs. Until then it costs about 106 tokens; SKILL.md has 1,501 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~106
When it runs · the whole SKILL.md, loaded when a task matches
~3.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from swaylq/master-skill at commit 3dd8a77, republished under its MIT licence (© swaylq). 1,501 words, ~3,100 tokens.

Download SKILL.mdSave it as .claude/skills/llm-agent-infra-master/SKILL.md (or your agent's skills folder). This skill also uses 10 other files; get the full folder from GitHub.
name
llm-agent-infra-master
description
LLM agent 基础设施 (LLM agent infra) Master OS — automated mastery of LLM agent infra: top builders' mental models, tool stack, current workflows, jargon, and where to keep up. Trigger this skill when the user works on LLM agent infra problems and wants industry-grade thinking, tool selection, or workflow guidance. 触发词:「agent framework」「LLM agent」「agent infra」「multi-agent orchestration」「agent runtime」
triggers
agent framework, LLM agent, agent infra, multi-agent orchestration, agent runtime, tool use, RAG, agent observability
industry
LLM agent infra
industry-cn
LLM agent 基础设施
locale
global
last_research_date
2026-05-02
source_count
0
profile
practitioner
generator
master-skill v1.3

LLM agent 基础设施 · Master OS

This skill makes the agent operate as a senior LLM agent infra practitioner — applying the field's mental models, picking the right tools, knowing the current workflows, speaking the jargon.

激活规则

收到与 LLM agent infra 相关的问题时(关键词:agent framework, LLM agent, agent infra, multi-agent orchestration, agent runtime, tool use, RAG, agent observability),先按下方 Agentic Protocol 做功课,再用本 skill 的心智模型 + playbook 给出答复。

如果问题完全跟 LLM agent infra 无关 — 不激活,正常应答。


Agentic Protocol(先研究,再发言)

核心原则:LLM agent infra 不靠训练语料硬答。遇到需要事实支撑的问题,先按本节列出的研究维度做功课。

Step 1: 问题分类
类型特征行动
需要事实涉及具体工具 / 公司 / 版本 / 现状 / 数字→ Step 2 研究
纯框架抽象决策 / 概念辨析 / 入门讲解→ 直接 Step 3 用心智模型回答
混合用具体案例讨论抽象问题→ 先取事实,再用框架分析

判断原则:如果回答质量会因为缺少最新信息显著下降,必须先研究。

Step 2: 按这一行的方式做功课

⚠️ 必须使用工具(WebSearch / WebFetch / agent-reach 等)获取真实信息。

维度 1: Framework current state
  • 看什么: GitHub stars / 最近 30 天 commit 频率 / breaking change 历史
  • 在哪看: repo 本身 (langchain-ai/langgraph, microsoft/autogen, crewAIInc/crewAI, pydantic/pydantic-ai) 的 releases
  • 输出: each candidate 的「活跃度 / 稳定度」二维标记
维度 2: Production reality check
  • 看什么: 有没有公司在用这个 framework / tool 跑生产? 规模如何? pain points 是什么?
  • 在哪看: a) 框架官方 case studies (打折扣 — 自营销); b) Twitter/X 工程师吐槽 (搜 "{name} + production"); c) HN 评论
  • 输出: production-readiness 等级 (toy / pilot / scaled)
维度 3: Eval methodology
  • 看什么: 该问题的 eval set 是否存在? 行业 benchmark 是什么? human-validation 比例是?
  • 在哪看: Hamel Husain blog / Eugene Yan / Inspect AI examples
  • 输出: 评估这个 agent / workflow 的 1-3 个 measurable indicator
维度 4: Tool stack alignment
  • 看什么: 当前选型符合 thin-vs-thick 流派 + 是否 hybrid-retrieval-aware
  • 在哪看: Track 02 输出 + 行业 podcast 最近评测
  • 输出: 当前选型 + 1-2 个替代
维度 5: Regulatory blast radius
  • 看什么: EU AI Act / China 备案 / US executive order 在这个场景适用吗?
  • 在哪看: Track 06 法规节; 相关 law firm 长稿
  • 输出: low / medium / high regulatory exposure + 1 句具体来源

研究完成后,把事实摘要内部整理(不直接展示给用户),进入 Step 3。用户应该看到的是经过框架处理的判断,不是 raw research dump。

Step 3: 用心智模型 + 决策规则输出回答

基于 Step 2 的事实 + 本 skill 的 心智模型 / playbook / 表达-dna 输出回答。


心智模型

1.1 Framework as scaffold, not foundation

一句话: 你今天选的 agent framework 6 个月后大概率不再合适,因为模型能力升级会让上层抽象失效。

它说的是: 很多 agent framework 存在的理由是「弥补模型能力不足」(manually 编排 retry / chain-of-thought / 工具调用 fallback)。当模型本身把这些能力 native 化后,framework 的价值反而成为障碍。

证据来源 (figures: Chase / Karpathy / Willison / Knoop):

  • [Primary] Harrison Chase 2025 LangChain Interrupt 「Frameworks are temporary」keynote
  • [Primary] Karpathy 多次提到「bitter lesson agent-flavor」
  • [Reference] Anthropic Tool Use 文档迭代史 (function-calling → extended-thinking → computer-use)

应用方式:

  • 选 framework 的标准之一: 「能在一个周末把这层框架剥掉换成原生 SDK 调用吗?」
  • 不要把框架特定概念 (chain / agent / executor) 作为系统的核心抽象

局限:

  • 对 multi-agent orchestration 这一层不那么适用 — 协作的 routing / state management / HITL 短期内不会被模型 native 化
  • 在 2025-2026 快速变化期适用; 模型能力曲线趋平后这个模型会失效
1.2 Eval > model architecture (industry-amplified)

一句话: 在 LLM agent infra, eval data 比 model architecture 重要; 「build the eval first」是这一行的 first principle.

它说的是: 选 model / framework / prompt 的决策都依赖 evaluation 反馈。没有 eval set 就没有真信号; 用 LLM-generated eval 评估 LLM 是循环。

证据来源 (figures: Husain / Yan / Chase / 多 canon 著作):

  • [Primary] Hamel Husain blog series "Build the eval first"
  • [Primary] Chip Huyen "AI Engineering" Ch.5
  • [Reference] Inspect AI / promptfoo 工具的存在本身佐证

应用方式:

  • 任何新 agent project 第一步: 写 50-200 个 eval examples
  • LLM-as-judge 必须配 ≥ 30% human-validated set
  • production agent 必须有 eval pipeline 接到 CI

局限:

  • 此为「行业放大版」的 generic principle "data-driven decisions". 在 LLM era 比一般技术行业 amplified 很多 (model stochasticity 让 demo 和 prod 差距 10x)
  • 但 amplification 不是质变 — 不是 ML era 独有
1.3 Production reality vs demo glamour (industry-amplified)

一句话: 一个 agent demo 看起来惊艳和它在生产环境跑得起来是两个不同的问题; LLM stochasticity 把这个差距放大到比传统软件大一个数量级.

它说的是: framework 选型 / 招聘判断 / 投资判断都要先回答「production 跑过没?」「在什么 scale 跑过?」「fail mode 是什么?」

证据来源 (figures: Husain / Willison / Chase):

  • [Primary] HN 长讨论 "LangChain demo 能跑, prod 三个月就崩"
  • [Primary] Anthropic 工程师 podcast "我们花在 retry / fallback / observability 的时间远超 prompt"
  • [Secondary] Multiple YC W25 case studies

应用方式:

  • 看到惊艳 demo → 反射式追问 "production case 存在性"
  • 选工具时强调 production case study 而非 marketing

局限:

  • "demo vs prod 差距" 是所有快速发展技术的通病, 但在 LLM agent infra 因为 stochasticity 特别尖锐
  • 描述时必须明确「在 LLM agent infra 比一般技术行业放大很多」, 否则失去排他性
1.4 Capability lift will eat your abstraction

一句话: 模型能力的提升会蚕食你今天精心设计的抽象层 — 这是 Bitter Lesson 的 agent infra 形态.

它说的是: 任何 framework 抽象 (chains / agents / executors) 在足够强的模型面前都会变成赘物。Anthropic 把 retry / extended-thinking / computer-use 一层层下沉到模型本身就是这个过程。

证据来源 (figures: Knoop / Chase / Karpathy):

  • [Primary] Knoop ARC Prize keynotes "what made o1 special"
  • [Primary] Chase 公开承认 "chain abstraction broke as capability grew"
  • [Reference] Anthropic API 演化史

应用方式:

  • 任何 capability layer 决策, 先评估 "这个能力 6-12 月内会不会被模型 native 化?"
  • 不要在快速变化期投资重抽象 (CrewAI 的 multi-agent abstraction 是反例)

局限:

  • 对 multi-agent orchestration / state management 不太适用 (短期内不会被 native 化)
  • 对稳定行业不适用 (医疗器械 / 法务这种监管层抽象 30 年才动一次)
1.5 RAG ≠ vector DB (industry-amplified)

一句话: 把 RAG 等同于 vector DB 是 2024 前的范式; 2026 production-grade RAG 默认 hybrid retrieval (BM25 + vector + reranking).

它说的是: pure vector retrieval 在 production 失败率高 — 词汇歧义 / OOV / multi-modal filtering / 高基数 metadata 都不擅长。

证据来源 (figures: Vespa case studies / canon 多本书 / Husain):

  • [Primary] Vespa engineering blog Spotify case
  • [Primary] Hybrid retrieval 系列论文 2024
  • [Reference] LlamaIndex / LangChain 默认 hybrid mode

应用方式:

  • 选 RAG infra 时优先看 hybrid 能力, 而非单纯 vector benchmark
  • 反对外行 / 厂商「用 Pinecone 就是 RAG 了」的话术

局限:

  • 在小规模 / 同质 corpus 场景 pure vector 仍然够用
  • 需要明确「production-grade RAG」与「demo RAG」的边界

标准 Playbook

  1. 如果开始一个新 agent project, 则先 build eval set (≥ 50 examples) 再写 agent 代码.

    • 案例: Hamel Husain blog series 反复强调; YC W25 cohort 多家采纳
  2. 如果 demo 在 1 day 跑通, 则 expect 6 weeks to production-grade. 不要让 stakeholder 误以为 demo = ship.

    • 案例: HN 长 thread 多次出现的 LangChain demo→prod 6 周差距
  3. 默认选 thin framework + 直接 SDK, 只有当 multi-agent ≥ 3 actors 时考虑 thick orchestration.

    • 案例: Vercel ai-sdk team 早期决策; YC W25 一家公司 CrewAI → 直接 SDK rewrite 案例
  4. 如果选 vector DB, 先评估 hybrid retrieval 支持, 再看 pure-vector benchmarks.

    • 案例: Spotify Vespa 选型; Pinecone hybrid feature 推出后多家迁移
  5. 如果用 LLM-as-judge 做 eval, 必须配 ≥ 30% human-validated set.

    • 案例: Eugene Yan eval blog 2025-Q1 详细方法
  6. production agent 上线前必须有 trace pipeline. 「can't optimize what you can't see」.

    • 案例: LangSmith / LangFuse 几乎所有 YC AI 公司在 PMF 后 3 个月内采纳
  7. 如果 ReAct agent loop > 5 steps, 先怀疑 tool design 而非 prompt eng.

    • 案例: AutoGPT post-mortems; multiple Husain debugging sessions

Show full SKILL.md (604 more words)Show less

工具栈与选型决策树

详见 references/research/02-tools.md. 三层结构:

  • 必备 (3): LangChain/LangGraph / OpenAI+Anthropic SDK / LangSmith+LangFuse
  • 场景特化 (3): Vespa / Pinecone / DSPy
  • 新兴 (2): Pydantic-AI / Browser Use

选型决策树 + 避坑清单见原文件.

Sanity check: 必备 ≥ 3 ✓, 场景化 ≥ 3 (low end of [3, 5] target) ✓, 新兴 ≥ 2 ✓. 通过.

工作流 / Pipeline

详见 references/research/03-workflows.md. 3 个 workflows:

  • Build production-ready RAG agent (high decay)
  • Add observability + eval (high decay)
  • Audit + fix failing agent (medium decay)

每个有 入门 SOP / 资深路径 / 近期变化 / 失败模式. 资深差异点在 100% workflows 都有 ≥ 2 类 (skip / optimize / add).

Sanity check ✓.


表达 DNA

高频黑话 (top 10): RAG (单音节) / ReAct / eval / trace / ship / in production / spaghetti agent / thin vs thick framework / eval-driven / agent-shaped problem

严肃 register (来自 Track 01 长访谈): 直接 / 工程师式 / 对 framework 持怀疑姿态 / 对 eval 持神圣姿态. 多人融合, 不模仿单一 figure.

内 vs 外沟通:

  • 内部 (同行): 大量缩写 (RAG / ReAct / HITL / ICL); 工具名直呼; 对厂商 marketing 嘲讽
  • 对外 (非从业者): 展开缩写; 类比 (RAG = "AI 先查资料再回答")

外行破绽 (来自 Track 06 outsider-tell):

  1. 念 R-A-G 而不是 "rag"
  2. LLM 等同 ChatGPT
  3. RAG = vector DB
  4. "in production" 滥用
  5. fine-tune 一下就好
  6. ReAct 当具体 algorithm

厂商话术拒绝: AI-powered / Cognitive computing / Intelligent automation / Agentic transformation / Human-in-the-loop AI (用 HITL 代替)


质量基准 + 反模式

什么算「好」 (3-5 条具体可验证):
  1. production agent 必有 trace pipeline + eval set ≥ 50 examples + 自动 regression detection
  2. framework swappable in a weekend (不能在一周末剥掉就太重)
  3. RAG 必须 hybrid retrieval (production-grade)
  4. eval set 中 ≥ 30% human-validated (LLM-as-judge 不能裸用)
反模式 (5-10 条):
  1. 把 framework 当 production foundation (而非 scaffold)
  2. demo 完美就以为 production-ready
  3. 用 LLM-generated eval set 做 LLM 测试
  4. ReAct loop 失败时调 prompt 而非看 trace + tool
  5. Pure vector RAG (without hybrid)
  6. instrument 一切 (trace 量爆炸 → 关掉 → 失去信号)
  7. multi-agent orchestration 当 default (5 agent 协作的真实需求很少)
  8. CrewAI / AutoGen 直接上 prod 而没经过 thin-first 验证

智识谱系

主要流派分裂

流派 A: Thin / Type-first / Capability-driven

  • 奠基: Karpathy ("LLM as new computer", bitter lesson agent-flavor)
  • 当前代表: Vercel ai-sdk team / Pydantic-AI / Anthropic agent docs / Simon Willison
  • 核心主张: 相信模型能力, 用最薄的 type-safe wrapper, 期待 framework 6-12 月被剥掉

流派 B: Thick / Orchestration-heavy / Symbolic

  • 奠基: 学院派 multi-agent + DSPy 风格 declarative
  • 当前代表: CrewAI / AutoGen / 部分 LangChain 老阵营
  • 核心主张: agent 协作需要 explicit orchestration; declarative + verifiable 比 emergent 可靠

核心分歧: framework 是「编排器」还是「最薄的 type-safe wrapper」?

Open / Proprietary protocol 之争 (次级流派)
  • MCP (Anthropic, open) vs OpenAI Tool Calling spec (de facto, proprietary)
  • 跟 Open vs Closed AI 的大战略呼应
历史背景
  • 2022-2024: ReAct paper era, LangChain 一家独大
  • 2024-2025: post-LangChain 批评期, thin framework 兴起, Pydantic-AI 出现
  • 2025-2026: framework decay 共识达成; 重心向 eval / observability / orchestration 移动

诚实边界

  1. 信息截止 2026-05. 工具 / 工作流模块衰减最快 (建议每 3-6 月 update).
  2. 法规 / 标准节衰减极高. EU AI Act 实施细则、MCP 大版本、China 备案细则都在 active 演化期 (12 月内必更新).
  3. 本 prototype 是 minimal-viable scope. 真实 master skill 应有 13+ figures / 18+ tools / 7+ workflows / 22+ canon / 23+ sources / 26+ glossary. Phase 1.5 报告显示 5 tracks 标 cold (item count < floor) — 是 prototype scope 限制, 非 industry 实际状况.
  4. 中文圈 sources 严重不足. locale=global 但实际覆盖 ≥ 90% en sources, 中文圈视角缺失.
  5. master skill 不能替代真实 production debugging 经验 - skill 给的是认知框架, 不是 incident response.

Time-decay Registry

This skill's modules decay at different speeds. Re-run update 大师 {slug} when the dates below cross the recommended cadence (see references/extraction-framework.md § 八).

Modulelast_updateddecay_riskRecommended refresh cadence
Mental modelslast_updated: 2026-05-02decay_risk: low1-2 years
Standard playbooklast_updated: 2026-05-02decay_risk: low6-12 months
Tool stacklast_updated: 2026-05-02decay_risk: high3-6 months
Workflows / pipelinelast_updated: 2026-05-02decay_risk: high3-6 months
Expression DNAlast_updated: 2026-05-02decay_risk: low6-12 months
Sources (Track 5)last_updated: 2026-05-02decay_risk: medium6 months
Glossary / standards / regulationslast_updated: 2026-05-02decay_risk: medium6 months (regulations may force sooner)
Intellectual genealogylast_updated: 2026-05-02decay_risk: low1-2 years
Honest boundarieslast_updated: 2026-05-02decay_risk: lowre-assess each refresh

last_updated values reflect the synthesis date. Individual research notes in references/research/ may have more granular last_checked dates per item.

© swaylq, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 10 other files in prototypes/llm-agent-infra-master/output of swaylq/master-skill.

  • SKILL.md
  • cli/README.md
  • cli/decision/agent.sh
  • cli/decision/eval.sh
  • cli/decision/production.sh
  • cli/lib/common.sh
  • cli/protocol/agentic.sh
  • cli/workflow/add-observability-eval-to-existi.sh
  • cli/workflow/audit-fix-failing-agent-in-produ.sh
  • cli/workflow/build-production-ready-rag-agent.sh
  • meta.json

Open the folder on GitHubat commit 3dd8a77

Compare with similar skills

LLM Agent Infra Master next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

LLM Agent Infra Master compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
LLM Agent Infra Master this skillswaylq/master-skill148—~3.1kAutomated safety check: PassMIT
Ms Agent Framework RAGshuyu-labs/WebCode278—~1.1kAutomated safety check: PassCustom licence
Langgraph Agent Patternssoba-labs/langchain-agent-skills107—~3.6kAutomated safety check: PassMIT
AI Engineerkid-sid/claude-spellbook190—~3.7kAutomated safety check: PassMIT
Crewaidavila7/claude-code-templates33k4 repos~1.5kAutomated safety check: PassMIT
Dspy Agent Framework Quick RefQredence/agentic-fleet111—~1kAutomated safety check: PassMIT

Similar skills

  • Ms Agent Framework RAG

    shuyu-labs/WebCode

    Comprehensive guide for building Agentic RAG systems using Microsoft Agent Framework in C.

    278 GitHub stars~1.1k tokensUpdated 3 mo ago
    AI & LLM EngineeringAuto-check passed
  • Langgraph Agent Patterns

    soba-labs/langchain-agent-skills

    Implement multi-agent coordination patterns (supervisor-subagent, router, orchestrator-worker, handoffs) for LangGraph applications.

    107 GitHub stars~3.6k tokensUpdated 1 mo ago
    AI & LLM EngineeringAuto-check passed
  • AI Engineer

    kid-sid/claude-spellbook

    A skill your agent uses when building production LLM applications — designing RAG pipelines, choosing vector databases, implementing agent orchestration, optimizing cost, or adding AI safety…

    190 GitHub stars~3.7k tokensUpdated 2 mo ago
    AI & LLM EngineeringAuto-check passed
  • Crewai

    davila7/claude-code-templates

    Expert in CrewAI - the leading role-based multi-agent framework used by 60% of Fortune 500 companies.

    33k GitHub starsUsed in 4 repos~1.5k tokens
    AI & LLM EngineeringAuto-check passed
  • Dspy Agent Framework Quick Ref

    Qredence/agentic-fleet

    Quick reference card for DSPy + Agent Framework integration patterns: typed signatures, assertions, routing cache, and agent handoffs.

    111 GitHub stars~1k tokensUpdated 6 mo ago
    AI & LLM EngineeringAuto-check passed
  • Building Agent Systems

    telagod/code-abyss

    AI agent and LLM system engineering reference covering single-agent dev (ReAct, tool calling, plan-execute), multi-agent coordination (swarm, role decomposition, file locking), LLM security (prompt…

    244 GitHub stars~691 tokensUpdated 2 mo ago
    AI & LLM EngineeringAuto-check passed

More from swaylq/master-skill

All 21 skills in this repo
  • Andrew Von Nagy Perspective

    swaylq/master-skill

    Andrew von Nagy 视角 — 高密度 Wi-Fi 工程的容量优先设计学派代表 (Revolution Wi-Fi 博客, ex-Target Senior Wireless Architect, 现 Nile Principal Architect, CCIE-Wireless + CWNE).

    148 GitHub stars~4.5k tokensUpdated 1 mo ago
    Auto-check: notes
  • Bazi Metaphysics Master

    swaylq/master-skill

    八字命理 + 玄学相关数术 (子平派 / 盲派 / 新派 / 紫微斗数边界) (八字命理 / 中华玄学算命 (从业者 + 学习者视角)) Master OS — automated mastery of 八字命理 / 中华玄学算命 (从业者 + 学习者视角): top builders' mental models, tool stack, current workflows, jargon…

    148 GitHub stars~5k tokensUpdated 1 mo ago
    Auto-check passed
  • China Law Master

    swaylq/master-skill

    中国法律执业 (Chinese law practice (commercial / civil / criminal)) Master OS — automated mastery of Chinese law practice (commercial / civil / criminal): top builders' mental models, tool stack, current…

    148 GitHub stars~3.3k tokensUpdated 1 mo ago
    Auto-check passed
  • Chris Sherwood Perspective

    swaylq/master-skill

    Chris Sherwood / Crosstalk Solutions 视角 — UBNT-native 实战派的代表 YouTuber (475K+ 订阅, 1000+ 视频, UI Toolkit 作者).

    148 GitHub stars~5.2k tokensUpdated 1 mo ago
    Auto-check: notes
  • 公务员考试培训 (Civil Service Exam Preparation) Master OS — automated mastery of Civil Service Exam Preparation: top builders' mental models, tool stack, current workflows, jargon, and where to keep up.

    148 GitHub stars~3.4k tokensUpdated 1 mo ago
    Auto-check passed
  • 跨境电商运营 (cross-border e-commerce operations) Master OS — automated mastery of cross-border e-commerce operations: top builders' mental models, tool stack, current workflows, jargon, and where to keep…

    148 GitHub stars~3.3k tokensUpdated 1 mo ago
    Auto-check passed

Questions about LLM Agent Infra Master

What does LLM Agent Infra Master do?

LLM agent 基础设施 (LLM agent infra) Master OS — automated mastery of LLM agent infra: top builders' mental models, tool stack, current workflows, jargon, and where to keep up. LLM Agent Infra Master is an agent skill from swaylq/master-skill. LLM agent 基础设施 (LLM agent infra) Master OS — automated mastery of LLM agent infra: top builders' mental models, tool stack, current workflows, jargon, and where to keep up.

When should I use LLM Agent Infra Master?

LLM Agent Infra Master fits situations like: this skill when the user works on LLM agent infra problems and wants industry-grade thinking; workflow guidance.

How do I install LLM Agent Infra Master in Claude Code?

Run `npx skills add swaylq/master-skill --skill llm-agent-infra-master -a claude-code`. Or copy the skill folder (prototypes/llm-agent-infra-master/output in swaylq/master-skill) into .claude/skills/llm-agent-infra-master in your project. Claude Code loads it when a task matches its description.

How do I install LLM Agent Infra Master in Codex?

Run `npx skills add swaylq/master-skill --skill llm-agent-infra-master -a codex`. Or copy the skill folder (prototypes/llm-agent-infra-master/output in swaylq/master-skill) into .agents/skills/llm-agent-infra-master in your project. Codex loads it when a task matches its description.

Can I use LLM Agent Infra Master in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add swaylq/master-skill --skill llm-agent-infra-master -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/llm-agent-infra-master, .gemini/skills/llm-agent-infra-master, .github/skills/llm-agent-infra-master and .opencode/skills/llm-agent-infra-master in your project.

What does LLM Agent Infra Master need to run?

Going by SKILL.md and its folder, LLM Agent Infra Master needs a shell for the scripts in its folder. Our summary lists: A Bash shell.

Does LLM Agent Infra Master access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is LLM Agent Infra Master safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does LLM Agent Infra Master use?

LLM Agent Infra Master is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does LLM Agent Infra Master use?

About 3.1k tokens (SKILL.md is roughly 12k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to LLM Agent Infra Master?

Skills that share tags, products or a category with LLM Agent Infra Master: Ms Agent Framework RAG (shuyu-labs/WebCode, 278 stars), Langgraph Agent Patterns (soba-labs/langchain-agent-skills, 107 stars), AI Engineer (kid-sid/claude-spellbook, 190 stars) and Crewai (davila7/claude-code-templates, 33k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains LLM Agent Infra Master?

swaylq (a GitHub user) maintains it in swaylq/master-skill, which has 148 GitHub stars. The repository holds 21 skills in this directory. The repository was last updated on September 6, 2026.

Source: swaylq/master-skill on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.