Coding prompts
Ready-to-paste prompts that help an AI write, check or improve code.
24 prompts. Each one is plain markdown you can paste into ChatGPT, Claude, Gemini or any AI chat, or install in Claude Code or Cursor. How it works
add-skills-from-theskillvault-prompt
A prompt that makes your AI coding agent search TheSkillVault for a ready-made skill before it starts a task, through the MCP server or the plain REST API, then load or install the best match. Use when the user asks to find a skill, wants their agent to look up skills automatically, or asks is there a skill for this task.
Not yet model-tested; testing Jan 2027Coding promptsanalyse-error-logs-prompt
A prompt that reads server, app or build logs and finds what matters - the first real failure, repeating patterns, spikes and timing, the probable cause with the log lines that prove it, and the next checks to run - without guessing beyond the data. Use when the user pastes logs, asks why a server crashed, or cannot spot what failed in a long log.
Not yet model-tested; testing Jan 2027Coding promptsbrainstorm-idea-into-design-prompt
A prompt that turns a rough idea into an agreed design before any code - the AI reads the project, checks what you want and why, one point at a time, offers two or three approaches with trade-offs, then writes a short design doc for you to approve. Use when you have a vague idea, want to brainstorm, or want options before building.
Not yet model-tested; testing Jan 2027Coding promptsbuild-throwaway-prototype-prompt
A prompt that has the AI build a throwaway prototype to answer one design question fast - a click-through HTML demo of the logic, or 3 radically different UI variants on one page with a switcher bar. Use when the user wants to prototype an idea, try a few layouts, check if a state model feels right, or mock something up before building it properly.
Not yet model-tested; testing Jan 2027Coding promptsbuild-with-parallel-subagents-prompt
A prompt that has the lead AI agent split work into independent tasks, send each to a fresh sub-agent with a focused brief, run them in parallel, then check for conflicts and run the full test suite. Use when the user has several unrelated bugs or failing tests, wants subagent-driven development, or asks to run parallel agents on a plan.
Not yet model-tested; testing Jan 2027Coding promptsexecute-plan-step-by-step-prompt
A prompt that has an AI coding agent carry out a written implementation plan one task at a time - test first, run the checks after every step, keep a short log of decisions, and stop only for risky or irreversible actions. Use when you already have a plan and want an agent to execute it one task at a time without drifting.
Not yet model-tested; testing Jan 2027Coding promptsexplain-code-prompt
A prompt that explains a piece of code, a script, a config file or tool output in plain English - what it does, a walk through of how it flows step by step, what the settings mean, any risks it carries - and answers your specific question about it. Use when you are handed code you do not understand, inherit a project, or want a walkthrough of what something does.
Not yet model-tested; testing Jan 2027Coding promptsmodel-domain-and-glossary-prompt
A prompt that sharpens the words your project uses - the AI challenges vague or clashing terms, tests them with edge-case scenarios, checks the code agrees, and records agreed terms in a GLOSSARY.md and big decisions as short ADRs. Use when names in the code are confusing, one thing has several names (customer, user, account), or you want a glossary.
Not yet model-tested; testing Jan 2027Coding promptsrespond-to-code-review-prompt
A prompt that makes an AI coding agent handle review feedback properly - restate each comment, check it against the real code, ask about anything unclear before changing a line, push back with reasons when a suggestion is wrong, then fix one item at a time. Use when you paste reviewer comments into an AI, or the agent agrees with every comment without checking.
Not yet model-tested; testing Jan 2027Coding promptsreview-code-prompt
A prompt for a thorough code review of a diff or pull request covering bugs, security, edge cases, performance and readability, with findings ranked by severity and concrete fixes. Use when the user wants code reviewed, a PR checked, a second opinion on changes, or to find bugs before merging.
Not yet model-tested; testing Jan 2027Coding promptsreview-codebase-architecture-prompt
A prompt that has the AI scan your codebase for shallow modules and tangled seams, rank "deepening" refactors that make code simpler to test and navigate, then grill you on the one you pick. Use when the user wants to improve codebase architecture, untangle messy code, plan a refactor, or make a repo easier for AI agents to work in.
Not yet model-tested; testing Jan 2027Coding promptsreview-plan-grill-me-prompt
A "grill me" prompt that makes the AI interview you round by round about a plan, design or idea, with a recommended answer for every question, until nothing is left assumed. Use when the user says grill me, wants their plan stress-tested, asks to be challenged before coding, or wants requirements nailed down first.
Not yet model-tested; testing Jan 2027Coding promptsrun-session-retro-prompt
A prompt for a short retrospective after an AI coding session - review what the agent struggled with and suggest fixes to its environment, such as navigation hints, automated checks, review rules, a leaner AGENTS.md and cheaper tool use, ranked by impact. Use when an agent session went badly, kept repeating mistakes, or you want future runs to go smoother.
Not yet model-tested; testing Jan 2027Coding promptswrite-agent-skill-prompt
A prompt that has your AI agent turn a task you keep repeating into a reusable SKILL.md agent skill - clear name, a description with Use when triggers, short numbered steps, gotchas and a done check - then test that it loads. Use when the user wants to create a skill, write a SKILL.md, save a workflow for Claude Code or Cursor, or improve an existing skill.
Not yet model-tested; testing Jan 2027Coding promptswrite-agents-md-prompt
A prompt that has your AI agent write or tidy an AGENTS.md project rules file - real build, test and lint commands, code style, folder map and do-not-touch rules - so every coding agent starts with the same context. Use when the user wants an AGENTS.md, CLAUDE.md or .cursor/rules file, says the AI keeps forgetting project conventions, or onboards an agent to a repo.
Not yet model-tested; testing Jan 2027Coding promptswrite-commit-message-prompt
A prompt that turns your staged changes or a short description of them into a clear Conventional Commits message with a type, scope, imperative subject, a why-focused body and breaking-change footer. Use when the user wants a commit message written, asks how to describe their changes, or follows conventional commits.
Not yet model-tested; testing Jan 2027Coding promptswrite-design-doc-prompt
A prompt that writes a technical system design document before a big build - goals, context, a C4-style context and container view with Mermaid diagrams, options considered, the chosen design, security and data risks, and open questions with assumptions. Use when the user plans a new system or major feature, needs a design doc or RFC, or wants architecture written down.
Not yet model-tested; testing Jan 2027Coding promptswrite-plan-and-tickets-prompt
A prompt that breaks a plan or spec into small tracer-bullet tickets, each a thin end-to-end slice with its blockers listed, then has the AI work through them one at a time with tests and commits. Use when the user has a spec to turn into tickets, wants a feature split into tasks, asks for a plan-then-execute workflow, or wants issues an agent can pick up.
Not yet model-tested; testing Jan 2027Coding promptswrite-pull-request-description-prompt
A prompt that writes a pull request description reviewers can act on fast - a small diagram or tree of what changed, before-and-after evidence such as test output or screenshots, and a merge-danger note on how reversible the change is and what it could affect. Use when the user opens a PR, asks for a PR body, or wants reviewers to understand a change quickly.
Not yet model-tested; testing Jan 2027Coding promptswrite-readme-prompt
A prompt that writes a clear README.md for a project from its code - what it does, a quick start that actually runs, configuration, usage examples and where to get help - in plain GitHub Markdown without emoji clutter or invented features. Use when a repo has no README, the README is out of date, or before making a project public.
Not yet model-tested; testing Jan 2027Coding promptswrite-regex-prompt
A prompt that writes a regular expression from a plain-English description for your language, then explains it piece by piece and tests it against strings that should and should not match, flagging slow or risky patterns. Use when the user needs a regex, asks to match emails, phone numbers or URLs, or a regex is not matching.
Not yet model-tested; testing Jan 2027Coding promptswrite-session-handoff-prompt
A prompt that compacts a long AI coding session into a short handoff document a fresh agent can pick up from, with links instead of copies and secrets redacted. Use when the chat is getting too long, context is running out, you are switching agents or tools, or you want to pick the work up later in a new session.
Not yet model-tested; testing Jan 2027Coding promptswrite-sql-from-plain-english-prompt
A prompt that writes a SQL query from a plain-English question - a correct, readable query for your database and schema - explicit columns, clear joins and filters, read-only by default - and explains what it returns and what to double-check. Use when the user needs a SQL query written, asks how to get some data out of their database, or does not know SQL.
Not yet model-tested; testing Jan 2027Coding promptswrite-terse-replies-prompt
A prompt that switches your AI agent into a terse reply mode - no filler, no repeated recaps, short fragments - to cut output tokens and reading time while keeping code, commands and errors exact. Use when the user says the AI is too wordy, wants shorter answers, wants to save tokens or cost, or asks for a caveman-style brief mode.
Not yet model-tested; testing Jan 2027