ai-engineering

ai-engineering is a skill for Claude Code, Codex from cass-2003/local-workflow-skill. It costs 130 tokens per session (6,976 once invoked), scanned A, original, MIT.

An engineering guide for building applications that use large language models, including chatbots, agents, search over documents, and tool calls.

In plain words
What is it for?
Designing, testing, monitoring, securing, and releasing AI features that use models, retrieved documents, tools, or multiple input types.
Why use it?
It helps teams control unreliable answers, security risks, changing model versions, response speed, and usage cost.

Skill for Claude CodeCodex

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.

agentmods
npx agentmods add skills/cass-2003/local-workflow-skill/ai-engineering
Any agent
npx skills add cass-2003/local-workflow-skill --skill ai-engineering
Clone the repo
git clone --depth 1 https://github.com/cass-2003/local-workflow-skill

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.

agentmods badge for ai-engineering

README.md
[![agentmods](https://agentmods.dev/badge/skills/cass-2003/local-workflow-skill/ai-engineering.svg)](https://agentmods.dev/skills/cass-2003/local-workflow-skill/ai-engineering)
Your own site
<a href="https://agentmods.dev/skills/cass-2003/local-workflow-skill/ai-engineering"><img src="https://agentmods.dev/badge/skills/cass-2003/local-workflow-skill/ai-engineering.svg" alt="Measured on agentmods" height="20"></a>
Per session 130 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 6,976 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00130 $0.06976
Opus 5 $0.00065 $0.03488
Sonnet 5 $0.00026 $0.01395
Haiku 4.5 $0.00013 $0.00698

Measured 4d ago against content hash 7cc61e0acddc, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

ai-engineering 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 4d 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/ai-automation/community/ai-engineering/SKILL.md · 212 lines

How it starts

The opening of the file, as written. The whole thing — 212 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AI 工程

AI 工程(ai-engineering,兼容 slug: aie)负责本技能描述范围内的定位、执行、验证和交接边界;旧短 slug 仅作兼容 alias/URL 主键,不作为规范技能名。

定位:把 AI / LLM 应用从“能跑 demo”收敛到可控、可测、可观测、可回滚、成本可接受、安全可上线。 铁律:没有模型、数据、工具、证据四类信息,不得宣称已优化;没有 eval / golden dataset / trace / 线上指标,只能标“需验证”。

开发门禁

  • 需求门禁:先写任务边界、目标用户、输入来源、输出用途、错误代价、不可接受行为、人工接管点和上线后业务指标;需求不清时不改 prompt 代替产品判断。
  • 指标门禁:每个 AI 功能至少定义质量、可靠性、安全、延迟、成本五类指标;没有基线、目标值和回滚阈值,不得宣称可上线。
  • 版本门禁:prompt、system/developer message、tool schema、RAG corpus、embedding model、rerank、eval dataset、guardrail、模型参数都要有版本号和变更记录。
  • 数据门禁:eval、日志、trace、prompt cache、Agent memory、向量库 ingestion 都要标注数据来源、权限、PII 状态、保留期、删除同步和训练/二次使用许可。
  • 实验门禁:模型选择必须给候选、证据、失败样例和取舍;只写“换更强模型”“调温度”“加一句 prompt”不是工程方案。
  • 发布门禁:灰度、A/B、kill switch、fallback、预算熔断、告警、回滚包、旧版本对照和负责人缺一项,只能发布到受控实验环境。

快速总则

  • 模型:记录 provider、model、version/date、region、context window、reasoning/vision/audio/tool 支持、限流、SLA、退役时间;Claude、OpenAI、Gemini、开源模型不可只按“聪明程度”替换。
  • 数据:用户输入、网页、文件、RAG 文档、工具结果、MCP resource、Agent memory 都是不可信数据;必须处理 PII、权限、来源、许可/版权、地域、保留期、删除同步、日志脱敏和二次使用授权。
  • 工具:tool use / function calling / MCP tool 只给最小白名单、强 schema、幂等、超时、审计;写操作必须 preview、权限校验、dry-run 或人工确认。
  • 证据:prompt、system prompt、RAG、embedding、向量库、chunking、rerank、guardrails、模型迁移都要绑定样例、指标、trace、测试命令或线上看板。
  • 验收:上线前必须定义任务级成功标准、失败代价、人工接管条件、业务 KPI/用户体验指标;只给离线准确率而无线上口径,结论只能写“需灰度验证”。
  • 输出:结构化输出优先 JSON schema / constrained decoding / provider 原生 structured output;解析失败必须有重试、降级、拒答或人工兜底。
  • 安全:prompt injection、防泄密、越权检索、跨租户缓存命中、工具误调用是上线门禁,不是上线后优化。
  • 红队:高风险链路必须覆盖 jailbreak、prompt injection、RAG poisoning、越权检索、工具参数 fuzz、数据外泄 canary、恶意文件和滥用输出评测;未跑红队不得宣称安全上线。
  • 观测:生产链路至少记录 request_id、tenant/user、provider、model、prompt_version、rag_version、tool、token、latency、cost、finish_reason、cache_hit、error_code;推荐 OpenTelemetry trace 串联。
  • 成本:优化前先拆 input/output token、reasoning token、embedding、rerank、tool roundtrip、stream 首 token、重试和缓存命中率,禁只看单次调用价格。
  • 回滚:模型、prompt、RAG 索引、tool schema、guardrail 和路由策略必须能独立回滚;回滚后要清理错误缓存、队列、索引别名和正在运行的 Agent 任务。

场景执行卡

0. 需求 / 指标 / 方案定义

  • 先查:用户任务、不可做事项、输出消费者、上下游系统、SLA、成本预算、合规边界、人工审核责任和已有失败样例。
  • 必做:把“成功”拆成可测指标;定义离线 eval 通过线、线上灰度阈值、投诉/拦截口径、人工接管和停止条件。
  • 方案:说明模型选择证据、prompt/tool/RAG 版本、数据集来源、fallback、缓存策略、观测字段、上线批次和回滚方案。
  • 验证:需求样例、负样例、边界长度、权限样例、旧 bug replay、成本延迟预算、数据脱敏检查。
  • 易漏:没定义失败代价就优化指标;只优化离线分数,忽略线上投诉、人工复核和业务转化。

Read the full file on GitHub · 212 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.

  1. 4d ago First seen · 212 lines · 130 tokens per session scan A 7cc61e0acddc

Subscribe to this mod's changes

ai-engineering is a skill published in the GitHub repository cass-2003/local-workflow-skill (12 stars, last pushed 1mo ago), licensed MIT. It adds 130 tokens to every session and 6,976 once invoked, about $0.0006 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.

Related

Other skills, from other repositories

dspy

DSPy: declarative LM programs, auto-optimize prompts, RAG.

mateaix/mateclaw · 19 tokens

data-flywheel

Use this skill when a user wants to turn repeated human-approved agent work across Claude Code, Hermes, OpenClaw, Codex, Cursor, or custom .agent/ loops into local artifacts for retrieval, evals, prompt shrinking, and optional future open-weight model/adapters.

codejunkie99/agentic-stack · 4 tokens

llm-application-dev

Building applications with Large Language Models - prompt engineering, RAG patterns, and LLM integration. Use for AI-powered features, chatbots, or LLM-based automation.

MoizIbnYousaf/Ai-Agent-Skills · 40 tokens

paranoia-ai-system-evolver

用于升级 AI 系统、agent workflow、Codex skill、prompt、memory、RAG、tool routing、schema、eval set 或 feedback loop;也用于把 AI 工作单从指令单升级为意图单,并对研究、检索、测试和 AI 对话做 VOI 决策门审计。需要 Intent Work Order、WOOP 任务准入、决策对象、VOI/EVPI/EVSI、UL(Uncertainty Ladder,不确定性阶梯)、OODA、eval、Human Gate、versioning 与 rollback 的受控演化时使用。Use when controlled AI system…

DY-2026/GameDesignOS · 147 tokens

ai-features

Use this skill when the user needs to add AI-powered features to their SaaS product, integrate LLM APIs, build AI assistants, implement RAG, or use AI to differentiate their product. Covers API selection, prompt engineering for product features, cost management, and building AI features that non-technical founders can…

whawkinsiv/solo-founder-skills · 70 tokens

reviewing-ai-papers

Analyzes an AI/ML publication — paper, preprint, article, technical blog post — and extracts what an enterprise AI engineer should do about it. Use when someone supplies a URL or document on RAG, embeddings, fine-tuning, prompt engineering, agents, or LLM deployment and asks "review this paper", "what do you make of…

oaustegard/claude-skills · 105 tokens