42 free prompts · all verified
Claude Code prompts
Framed as reusable rules/skills templates, not one-off copy-paste prompts — practitioners have shifted toward CLAUDE.md-style context files over flat prompts, and a library that still treats this as "type a prompt" is already behind.
Every prompt lists the exact tool and version it was tested against.
Write a CLAUDE.md file that still holds up after fifty sessionsA structured CLAUDE.md authoring prompt that separates durable project facts from session-specific noise, so the file Claude Code auto-loads every session stays accurate instead of silently rotting into stale advice.Claude Code2026-07-21Author a custom slash command Claude Code will run the same way every timeA prompt that generates a well-formed .claude/commands/*.md custom slash command — frontmatter, argument handling, and an explicit output contract — so a repeated workflow becomes one typed command instead of a differently-worded prompt each time.Claude Code2026-07-22Define a specialized Claude Code subagent instead of overloading the main threadA prompt that produces a complete .claude/agents/*.md subagent definition — scoped tools, a narrow system prompt, and a model choice — for a repeated task that deserves its own isolated context rather than living in the main conversation.Claude Code2026-07-23Write a PreToolUse hook that blocks a dangerous command before Claude Code runs itA prompt that generates a Claude Code PreToolUse hook — matcher pattern, hook script contract, and exit-code semantics — that mechanically blocks a specific risky action instead of relying on a CLAUDE.md rule the agent could still reason its way past.Claude Code2026-07-24Configure a Claude Code permissions allowlist that matches how the team actually worksA prompt that audits a described workflow and produces a settings.json permissions block — allow, ask, and deny lists — scoped to exactly the tool patterns that workflow needs, instead of a blanket approval that trades away the point of having a permission system.Claude Code2026-07-25Force a written plan before Claude Code touches a multi-file featureA plan-first prompt that blocks Claude Code from creating, editing, or deleting any file until it has produced and gotten approval on a full change plan, mirroring the discipline of Plan Mode itself but making it a standing requirement rather than a manual toggle.Claude CodeCursor2026-07-27Run Claude Code's structured security review discipline before every mergeA closed-taxonomy security review prompt that separates flagging a finding from fixing it, mirroring the mechanics of Claude Code's built-in security review command, so a vulnerability finding produces a human decision rather than a silent autonomous patch.Claude Code2026-07-28Turn Claude Code into a consistent code reviewer instead of a vibes checkA fixed six-category code review rubric with a forced pass, concern, or blocking rating per category and a single final verdict, so every pull request gets the same systematic pass instead of whatever the model happens to notice first.Claude CodeCursor2026-07-29Lock Claude Code to the minimal fix and stop the drive-by refactorA constraint block for a bug report that pins Claude Code to exactly the acceptance criteria stated, with an explicit list of what counts as scope creep even when the extra change looks like an improvement.Claude CodeCursor2026-07-30Make Claude Code write tests that hunt for breakage, not confirm the happy pathA test-generation prompt that enumerates required edge-case categories and forces table-driven coverage, directly countering the documented happy-path bias of LLM-written tests.Claude CodeCursor2026-07-31Refactor with Claude Code without smuggling in a behavior changeA refactor prompt that makes preserving external behavior an explicit, checkable constraint, forces the existing test suite to pass unmodified as proof, and requires the diff to be explained hunk by hunk, not just dumped.Claude CodeCursor2026-08-01Turn a raw diff into a pull request description a reviewer can actually useA format-forced, audience-aware prompt that turns a raw diff into a structured pull request description written for a reviewer who has not opened the files yet, instead of a generic one-line summary.Claude CodeCursor2026-08-02Diagnose a flaky test's real cause before Claude Code touches itA diagnosis-first prompt that stops Claude Code from silently wrapping a flaky test in a retry or a longer timeout, and instead forces it to isolate the actual source of nondeterminism through repeated real test runs and git history.Claude Code2026-08-03Map a major dependency upgrade before Claude Code touches the lockfileA pre-upgrade research prompt that has Claude Code cross-reference every breaking change in a dependency's release notes against real usage sites in the repo, catching the silent behavior changes a live-fire upgrade-and-fix-what-breaks approach reliably misses.Claude CodeCursor2026-08-04Recover a Claude Code task correctly after a compaction or restartA session-recovery prompt that has Claude Code reconstruct exactly where a multi-step task stands from git state and any saved plan file, rather than trusting a compacted summary or its own fluent-sounding recollection of what already happened.Claude Code2026-08-05Get an unfamiliar codebase explained top-down before touching a single lineAn onboarding prompt that forces a layered explanation — entry point, architecture, data flow, then the specific traps — instead of a flat file-by-file dump that burns attention without building a mental model.Claude CodeCursor2026-08-06Script a Claude Code headless run that a CI pipeline can actually trustA prompt structured for Claude Code's non-interactive headless invocation — explicit output-format contract and exit-condition rules — so a CI step gets a parseable, deterministic result instead of conversational prose no script can act on.Claude Code2026-08-07Run two Claude Code sessions in parallel on the same repo without them collidingA worktree-setup prompt that has Claude Code create an isolated git worktree per parallel task before starting work, so two agentic sessions on the same repository never fight over the same files or an uncommitted change in the other session's way.Claude Code2026-08-08Build a custom Claude Code output style and statusline instead of living with the defaultsA prompt that produces both a custom output-style definition and a statusLine script, tuned to how a specific developer actually wants Claude Code to communicate and what they need visible at a glance in the terminal.Claude Code2026-08-09Orchestrate several Claude Code subagents on one task without collisionsA delegation prompt for a main Claude Code session coordinating multiple distinct subagents on one complex task — researcher, implementer, reviewer — with an explicit handoff contract between them, so parallel or sequential delegation does not produce duplicate work, conflicting edits, or a finding that silently never reaches the implementer.Claude Code2026-07-20Scope Claude Code's .mcp.json to the project instead of trusting global defaultsA configuration prompt for a project-level .mcp.json that names exactly which MCP servers this repository needs, keeps secrets out of a file that is often committed, and treats adding a new server as a reviewable trust decision rather than a default the whole team inherits silently.Claude Code2026-07-20Turn a raw production stack trace into a scoped root-cause hypothesis, not a guessAn incident-triage prompt that has Claude Code correlate a raw stack trace against recent deploys and error-frequency data before proposing a root cause, and separates an immediate mitigation from a permanent fix so the two are never accidentally conflated under time pressure.Claude Code2026-07-21Review a database migration for what it does to production before it runs thereA pre-execution review prompt that checks a migration script's locking behavior, backward compatibility during a rolling deploy, and reversibility against the actual table size and deploy strategy, instead of judging it purely on SQL syntax correctness.Claude Code2026-07-21Keep Claude Code inside one monorepo package instead of reaching across itA scope-lock prompt for monorepo work that names the exact package boundary in play, treats a shared-package edit as a decision requiring explicit surfacing rather than a silent side effect, and distinguishes a genuine cross-boundary need from a workaround for a problem that actually belongs inside the target package.Claude Code2026-07-22Fix a UI bug from a pasted screenshot instead of a vague text descriptionA visual-debugging prompt built around a pasted image of the broken UI, forcing Claude Code to name the specific visual discrepancy and locate the responsible component before editing, and to close the loop with a follow-up screenshot comparison instead of declaring the fix done from the diff alone.Claude Code2026-07-22Generate docs that match what the code actually does, not what it should doA documentation-generation prompt that derives every claim from real exported signatures, behavior, and tests rather than from the intent behind the code, and forces an explicit flag on any doc statement the reviewer cannot verify against the code itself.Claude Code2026-07-23Wire GitHub's claude-code-action without granting it more authority than the job needsA workflow-authoring prompt for anthropics/claude-code-action that scopes the triggering condition, the GITHUB_TOKEN permissions block, and the allowed-tools config to exactly what the automated job does, since an unattended GitHub Actions run has no ask step to fall back on if a permission is too broad.Claude CodeGitHub Actions2026-07-23Audit where a Claude Code session's cost actually went before setting a budgetA usage-audit prompt that breaks down a session's actual token and cost data by category — context reads, tool-result bloat, repeated file access — before proposing a concrete reduction lever and a going-forward budget, instead of guessing at savings from vague impressions of the session.Claude Code2026-07-24Use Claude Code's checkpoint and rewind deliberately, not as an afterthoughtA discipline prompt for sequencing a risky experimental change around Claude Code's checkpoint and rewind feature, distinguishing what a rewind actually restores — conversation and file state — from what it cannot undo, such as an external side effect already sent, so genuinely irreversible steps are never attempted before a checkpoint exists.Claude Code2026-07-25Structure a long Claude Code session with TodoWrite instead of losing track of what's leftA task-tracking discipline prompt that has Claude Code decompose a long task into a TodoWrite list at the start, keep exactly one item in progress at a time, and treat an item marked complete without stated verifiable evidence as not actually complete — closing the gap between looking organized and actually staying on track.Claude Code2026-07-26Reach for 'think hard' deliberately, not as a reflex on every requestA calibration prompt for when to actually invoke Claude Code's extended-thinking trigger words versus when they add latency and cost with no real benefit, tied to task reversibility and complexity rather than habit, and requiring the reasoning budget to be stated rather than left to the keyword alone.Claude Code2026-07-27Package a repeated procedure as a SKILL.md Claude Code discovers on its ownA skill-authoring prompt that produces a complete SKILL.md — trigger description, procedural body, and any bundled scripts or reference files — for a multi-step procedure the main agent should recognize and apply automatically, distinct from a slash command's explicit invocation and a subagent's isolated context.Claude Code2026-07-28Plan a repo-wide pattern migration as a codemod, not five hundred manual editsA large-scale transformation plan that defines the exact before-and-after pattern once, verifies it against a codemod tool on a sample before running it repo-wide, and stages rollout by directory — distinct from a single dependency's breaking-change migration, this is for a systemic pattern change across the whole codebase.Claude Code2026-07-29Freeze the interface first, implement against it, instead of discovering it mid-buildA contract-first development loop that has Claude Code define the full interface, type, and error-case shape for a feature before any implementation begins, requires explicit re-approval for any later contract change, and names every downstream consumer that would be affected if the frozen shape turns out to need adjusting.Claude Code2026-07-30Audit Claude Code's Bash permissions for the escape route a single rule missedA holistic security audit of an existing settings.json's full Bash permission surface — checking whether the union of individually reasonable allowed patterns can be composed into a forbidden action — rather than reviewing any single rule in isolation, plus a check of the actual sandbox mode restrictions in effect.Claude Code2026-07-31Use the Claude Code IDE extension's own diff view instead of working blind in a terminalA prompt tuned for the VS Code and JetBrains Claude Code extensions that leans on the IDE's automatically-shared selection, open file, and diagnostics context, routes review through the IDE's native inline diff instead of a pasted patch, and hands verification off to the IDE's own linter and test runner instead of re-implementing that check inside the prompt.Claude Code2026-08-01Wire a PostToolUse hook that auto-formats every edit instead of nagging in reviewA PostToolUse hook prompt for automatically running a formatter or linter after every Edit or Write call and feeding back only what actually changed — a mechanically distinct hook lifecycle from PreToolUse, since PostToolUse cannot block the action that already happened but can surface a correction as fresh context Claude acts on immediately in the same turn.Claude Code2026-08-02Run a notebook data analysis with Claude Code's NotebookEdit without losing outputsA cell-level notebook workflow prompt that has Claude Code edit a Jupyter notebook through NotebookEdit rather than a whole-file rewrite, preserving already-computed cell outputs that took real time to generate unless a cell's actual inputs changed, and requiring a stated hypothesis before an exploratory cell runs.Claude Code2026-08-03Write a MIGRATION.md ledger Claude Code re-reads every session of a multi-week migrationA persistent rules-and-ledger file for long-running code migrations — the kind that outlives any single Claude Code session — tracking what has been moved, what is frozen, and how to verify each step before touching the next file.Claude Code2026-08-08Brief Claude Code to port a codebase between languages without smuggling over the source language's idiomsA four-phase task brief for porting a module or service from one programming language to another, forcing an explicit idiom-mapping pass before any file gets touched so the result reads as native target-language code, not translated source code.Claude Code2026-08-10Get an architecture review from Claude Code that cites file and line, not a generic 'consider microservices' essayA tightly bounded prompt for a structured architecture review — layering violations, coupling hotspots, and a prioritized list of concrete findings, each backed by an actual file reference rather than abstract best-practice advice.Claude Code2026-08-12Set up a TECH_DEBT.md ledger Claude Code appends to instead of fixing things on the slyA rules file that turns technical debt from a vague, unprioritized complaint into a tracked, file-and-line-referenced ledger with effort and impact scored per item, so debt gets paid down deliberately instead of opportunistically rewritten mid-task.Claude Code2026-08-14
