Vibe Prompting
Prompts that actually ship
A working library of prompts I reach for daily, plus the workflow I use to turn a vague idea into something verified and reusable. Copy any prompt with one click.
The workflow
How a prompt goes from fuzzy ask to shipped. Your progress is saved locally.
Step 01
Frame
Turn a fuzzy ask into a verifiable goal before touching the model.
The library
- engineering
Senior Code Review Partner
A prompt that turns an LLM into a rigorous, kind senior reviewer focused on correctness, edge cases, and clear tradeoffs.
You are a senior software engineer doing a careful code review. Review the diff or code I provide with this priority order: 1. Correctness bugs and edge cases (null/empty/overflow/concurrency). 2. Security issues (injection, auth, secrets, unsafe deserialization). 3. Readability and naming that will confuse the next reader. 4. Simplifications — flag any code that could be meaningfully shorter. Rules: - Quote the exact lines you're commenting on. - For each issue, state severity (blocker / nit) and a concrete fix. - Do NOT rewrite unrelated code or restyle to your taste. - If something is genuinely fine, say nothing about it. - End with a one-line verdict: APPROVE, APPROVE WITH NITS, or REQUEST CHANGES. Here is the code: ``` <paste code here> ``` - workflow
Force a Spec Before Any Code
A planning prompt that makes the model surface assumptions and a verifiable plan before writing a single line — killing the over-eager-coder failure mode.
Before writing ANY code for the task below, produce a short spec: 1. Restate the task in one sentence. 2. List your explicit assumptions. Mark any you're unsure about with (?). 3. Note any ambiguity that has more than one reasonable interpretation. 4. Propose the simplest approach. If a simpler one exists than what I asked for, say so. 5. Give a step-by-step plan where every step has a verification check: Step -> verify: <how we'll know it worked> Then STOP and wait for my confirmation. Do not write code yet. Task: <describe the task here> - engineering
Debug by Hypotheses, Not Guesses
A debugging prompt that forces the model to rank hypotheses and design the cheapest experiment to discriminate between them before touching code.
Help me debug this. Work like a scientist, not a guesser. 1. Summarize the symptom in one line and what "fixed" would look like. 2. List 3-5 plausible root-cause hypotheses, ranked by likelihood. 3. For the top hypothesis, design the SINGLE cheapest experiment that would confirm or rule it out. Prefer a log line or a one-off check over a code change. 4. Tell me exactly what to run/observe and what each outcome would mean. 5. Wait for the result before proposing the next step. Do not propose a fix until a hypothesis is confirmed. Context: - Symptom: <what's happening> - Expected: <what should happen> - Relevant code/logs: <paste>