通过提示词优化性能
让人工智能找出运行缓慢的路径。
通过提示词优化性能 是 CoddyKit 上的免费 Vibe Coding 课时。 这是第 2 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 Vibe Coding 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Vibe Coding 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
Measure, Don't Guess
Performance work fails when it starts with hunches. The fastest way to slow yourself down is to optimize code that was never the bottleneck.
Your assistant is excellent at optimization, but only if you point it at real evidence. This lesson is about turning measurements into precise prompts.
Profile First
Before asking for speed, ask for a profile. Have the assistant instrument the slow path and tell you where time actually goes.
Give it the symptom and the constraint, and let it propose where to measure instead of guessing at fixes.
The /search endpoint takes about 2 seconds under load. Before optimizing anything, add profiling to break down where the time goes: database queries, serialization, and external calls. Tell me which segment dominates and propose where to focus.Hunt the N+1
The most common performance killer in a vibe-coded app is the N+1 query: a loop that fires one database call per item. It feels fine with ten rows and dies at ten thousand.
Ask the assistant to find these patterns and collapse them into a single batched query or a join.
Scan the data-access layer for N+1 query patterns where we loop over results and issue a query per item. Rewrite the worst offender to use a single batched query or a join, and show the before/after query count for a list of 100 items.Index the Right Columns
A query that scans the whole table is slow no matter how clean the code is. The fix is usually an index on the columns you filter and sort by.
Share the slow query and let the assistant read the execution plan and recommend a targeted index, not a blind one on every column.
Here is a slow query and its EXPLAIN output. Recommend the minimal set of indexes to remove the sequential scan, explain the trade-off on write performance, and give me the exact migration to add them.Cache What's Expensive
Some results are costly to compute and rarely change. Caching them trades a little staleness for a large speedup.
Ask for a cache with an explicit invalidation strategy, because a cache you cannot invalidate correctly becomes a bug factory.
Add a cache layer for the product-catalog response, which is expensive to build and changes a few times a day. Use a short TTL plus explicit invalidation when a product is updated. Explain exactly when stale data could appear and how long it lasts.Move Slow Work Off the Request
Sending an email, resizing an image, or calling a third-party API inside a request makes the user wait for work they don't need to see finish.
Have the assistant move that work into a background job so the response returns immediately and the task runs reliably with retries.
The signup request blocks while sending a welcome email and provisioning resources. Move that work into a background queue so the request returns immediately. Add retries with backoff and a dead-letter path for tasks that keep failing.Paginate and Stream
Returning ten thousand rows in one response is slow to build, slow to transfer, and slow to render. Products page through data instead.
Ask for cursor-based pagination for large lists and streaming for large payloads so memory stays flat as data grows.
The list endpoint returns the entire table in one response. Add cursor-based pagination with a sane default page size and a stable sort. Make sure memory usage stays constant regardless of total row count.Set a Budget
Optimization without a target never ends. Define a performance budget so you know when you're done and when you've regressed.
Express it concretely: a p95 latency, a payload size, or a query count, and ask the assistant to keep changes inside it.
Set a performance budget for the checkout flow: p95 latency under 300ms and no more than 5 database queries per request. Review the current flow against that budget and propose the smallest set of changes to meet it.Beware Premature Optimization
Not every slow thing is worth fixing. Optimizing a path that runs once a day wastes effort and adds complexity that bites you later.
Ask the assistant to weigh impact against risk, and to leave clear code alone when the gain is negligible.
Here are five optimization ideas from the profiler. Rank them by expected impact versus the complexity and risk each one adds. Recommend which to do now, which to defer, and which to skip because the gain is too small to justify.Verify the Win
An optimization isn't real until it's measured again. The same profile that found the problem should confirm it's gone, with no new regression elsewhere.
Have the assistant re-run the benchmark and report the delta honestly, including any path that got slower.
Re-run the benchmark on /search after the changes and compare to the baseline. Report p50 and p95 before and after, the query count delta, and flag any other endpoint that regressed as a side effect.Performance Is a Loop
Speed is never finished. As usage grows, new bottlenecks appear where there were none before.
The durable skill is the loop itself: measure, target the biggest cost, change, and re-measure. Keep your prompts anchored to data and the loop keeps paying off.
Quick Check
Test your understanding of prompting for performance.
Recap
Performance by prompt means feeding the assistant evidence, not guesses. Profile first, then target N+1 queries, missing indexes, expensive uncached work, and blocking tasks that belong in the background.
Set a budget, verify each win with the same benchmark, and treat optimization as a continuous loop driven by measurement.
常见问题解答
「通过提示词优化性能」课时是免费的吗？
是的 — 「通过提示词优化性能」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Vibe Coding 课程的其余内容，请升级到 CoddyKit PRO。 Vibe Coding 课程共包含 4 节课。
「通过提示词优化性能」这节课中我会学到什么？
让人工智能找出运行缓慢的路径。 你通过在浏览器中直接运行的动手代码来练习 Vibe Coding，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 Vibe Coding 需要有经验吗？
无需任何先前经验。CoddyKit 上的 Vibe Coding 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 2 节课，共 4 节。
「通过提示词优化性能」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 Vibe Coding 课中编写并运行代码吗？
能。每节 Vibe Coding 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。