4.1 KiB
4.1 KiB
LinkedIn Draft & Consultant Review: Stop Writing Agent Prompts for Deterministic Work
文档性质：从
blog.hoelee.com改写并在 NocoDB 排期的 LinkedIn 生产级动态。 发布状态：已通过外部顾问审阅 / NocoDB 已排期待发布排期波次：Wave 2 (纵深期) - 槽位 8 (拟发布日期: 2026-11-24 Tue) 源文章：https://blog.hoelee.com/posts/stop-writing-agent-prompts-for-deterministic-work/ 系统标识：stop-writing-agent-prompts-for-deterministic-work| 分类：notes字数：1959 chars (严格控制在 3,000 字符内)
一、 核心背景与铁律约束提示（供复核对照）
- 求职画像契合（Senior 全栈/DevOps/AI 架构师）：彻底脱敏“做生意的老板/小公司创始人”倾向；一切商业实战均转换为“生产系统架构、高可用交付与自动化运维”视角。
- 纯洁人设与零禁用词：绝对禁止
homelab,home lab,家用,小主机,玩具,折腾,玩票等词汇。必须使用the infrastructure I manage,my self-hosted production nodes,production host。 - 第一人称单数：统一使用单数
I/我，杜绝虚张声势的复数we。 - 数据绝对保真：所有数字、性能比率、吞吐提升、故障现象严禁凭空捏造。
- OPSEC 安全脱敏：无内部绝对路径（如
D:/dev），无真实私有内网 IP，公网 IP 必须转为文档保留地址。
二、 LinkedIn 拟发布动态文案（已按顾问意见修订）
One of my automated background jobs failed 29 times in a row without sending an alert.
The core check was trivial: query npm for the latest published version of a CLI tool, compare it with the currently installed version, and send a notification if a newer version exists.
Two strings and a version inequality.
Because I already had an autonomous agent scheduler running daily ops summaries, I lazily wired this check as an agent prompt task instead of writing a dedicated script. Every cycle, an LLM parsed the CLI output and decided whether to alert. It worked fine until an upstream model routing change occurred. The agent framework recorded a model mismatch, triggered a silent drift skip, and exited cleanly without error.
The fix was not a more defensive prompt. It was deleting the LLM from the loop entirely.
The rule I enforce across all automated pipelines now:
**If the exit condition is a deterministic comparison — a string, a number, a status code — putting a model in that loop only adds failure modes.**
There is zero judgment required in `currentVersion < latestVersion`. Wrapping it in an LLM adds:
• Network latency and provider downtime
• Model drift and prompt deprecation risks
• Non-deterministic output formatting
• Unnecessary token billing
I replaced the agent task with a deterministic 125-line Python script handling HTTP retries, state deduplication, and explicit exit codes. It has run cleanly ever since.
Of the eleven automated operations loops I manage today, eight contain zero AI models in their execution path. The remaining three use LLMs exclusively where human ambiguity cannot be expressed as code: ranking unstructured emails, evaluating qualitative risk, and flagging stalled project contexts.
Where in your workflows are you using prompt engineering where a deterministic script would be infinitely more resilient?
Write-up: https://blog.hoelee.com/posts/stop-writing-agent-prompts-for-deterministic-work/
三、 外部顾问审阅评估反馈单与归档记录
- 审阅结论：需微调（已更新修订稿）
- 1. 诊断与修改批注：防杠微调：交代最初挂在调度器背景与 125 行包含重试/去重/退避逻辑，立住工程克制力形象。
- 2. NocoDB 归档记录：
- Candidate Key:
cand:stop-writing-agent-prompts-for-deterministic-work(Status:APPROVED, Priority:HIGH) - Draft Key:
cand:stop-writing-agent-prompts-for-deterministic-work:v1(Status:APPROVED) - 拟定发布节奏：每周 2 篇（周二/周四），平稳推进 5 周。
- Candidate Key: