菜单
Use when designing an LLM application, deciding whether it needs a direct call, deterministic workflow, agent, multi-agent system, or framework, and defining model boundaries, state, structured output, streaming, failure handling, and human control.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
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/blackprince001/skills --skill ai-ml-build-llm-app-architecture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。
Always-on coding style and organization rules for naming, file shape, comments, types, module boundaries, pattern use, errors, and avoiding unnecessary abstraction across languages and tracks.
Use when setting up or modifying CI pipelines — the quality-gate sequence every change passes, shift-left ordering, feeding failures back to agents, and keeping the pipeline under 10 minutes.
Use when designing any public interface — REST/GraphQL endpoints, module contracts, component props — Hyrum's Law, contract-first, consistent error semantics, validate at boundaries, extend rather than break.
基于 SOC 职业分类
正在显示 SKILL.md
| name | ai-ml-build-llm-app-architecture |
| description | Use when designing an LLM application, deciding whether it needs a direct call, deterministic workflow, agent, multi-agent system, or framework, and defining model boundaries, state, structured output, streaming, failure handling, and human control. |
Apply core architecture, contracts, state/data, security, performance, and observability first.
Read references/source-notes-agent-building-guides.md
when evaluating agent complexity or frameworks.
Use this progression and stop at the first level that meets the eval and product target:
Use a framework only when its durable execution, state, tracing, checkpoint, provider, or governance capabilities save more complexity than its abstraction adds. Prototype the critical path without framework-specific domain logic and record an exit strategy.
Keep provider calls behind an application-owned interface. Version model configuration, prompts, tools, schemas, retrieval policy, and guardrails as one behavioral release. Validate every model output before application use; treat free text as presentation, never authority. Keep deterministic business rules, authorization, money, identity, and irreversible state transitions in code.
Define request identity, timeout, cancellation, retry classification, token/turn/tool budgets, idempotency, and degraded behavior. Retry transient provider failures with bounded backoff; do not retry a completed side effect. Stream only when partial output improves the experience and can be cancelled, moderated, and safely abandoned.
Separate conversation history, working state, durable memory, retrieved evidence, and audit state. Persist only what has a product purpose, retention rule, tenant boundary, and deletion path. Give every tool a risk rating. Require human confirmation for consequential, high-value, externally communicative, destructive, irreversible, or low-confidence actions; escalate on repeated failure or ambiguity.
The architecture record must show why AI and the selected autonomy level are necessary, the eval baseline that earned them, model/tool/data trust boundaries, state ownership, limits, human gates, failure and rollback behavior, cost/latency budgets, and production signals.
