cursor/plugins
63 skills52k installs160k starsGitHub
Install
npx skills add https://github.com/cursor/pluginsSkills in this repo
1Thermo Nuclear Code Quality ReviewRun an extremely strict maintainability review for abstraction quality, giant files, and spaghetti-condition growth. Use for a thermo-nuclear code quality review, thermonuclear review, deep code quality audit, or especially harsh maintainability review. --- name: thermo-nuclear-code-quality-review description: Run an extremely strict maintainability review for abstraction quality, giant files, and spaghetti-condition growth. Use for a thermo-nuclear code quality review, thermonuclear review, deep code quality audit, or especially harsh maintainability review. disable-model-invocation: true --- # Thermo-Nuclear Code Quality Review Use this skill for an unusually strict review focused on implementation quality, maintainability, abstraction quality, and codebase health. Above all, this skill should push the reviewer to be **ambitious** about code structure. Do not merely identify local cleanup opportunities. Actively search for "code judo" moves: restructurings that preserve behavior while making the implementation dramatically simpler, smaller, more direct, and more elegant. ## Core Prompt Start from this baseline: > Perform a deep code quality audit of the current branch's changes.8.9kinstalls2DeslopThe deslop skill from Cursor plugins cleans AI-introduced noise in pull request diffs against main. Focus areas include unnecessary comments inconsistent with file style, abnormal defensive try/catch on trusted paths, any casts used only to silence types, deeply nested logic that should use early returns, and other patterns unlike surrounding code. Guardrails require unchanged behavior unless fixing a clear bug, minimal focused edits over broad rewrites, and a concise 1-3 sentence final summary. Agents diff the branch against main, target slop introduced in the branch only, and align edits with local conventions per file. Use when users ask to deslop, remove AI slop, or clean up AI-generated code style in a PR.3kinstalls3Fix Merge ConflictsFix Merge Conflicts is a skill from cursor/plugins that runs when a branch has unresolved merge conflicts and needs a reliable path to a buildable state. The workflow detects conflicting files from git status and conflict markers, resolves each with minimal correctness-first edits, prefers preserving both sides when safe, and regenerates lockfiles with package managers instead of hand-editing. It then runs compile, lint, and relevant tests before staging resolved files and summarizing decisions. Developers reach for Fix Merge Conflicts after merges or rebases leave conflict markers blocking CI. The skill finalizes resolution non-interactively inside the editor session.1.6kinstalls4Make Pr Easy To Reviewmake-pr-easy-to-review is a Cursor plugin skill that makes pull requests easier for teammates to review. The workflow resolves the target PR from a URL or current branch, inspects commits, diff size, changed paths, generated files, and the existing description, then fixes reviewability issues such as noisy commits, stale summaries, and missing risk context. The default goal is reviewability without behavior changes—annotating diffs and tidying history rather than refactoring logic. Developers invoke it for requests like "make this easy to review", "tidy this PR", "clean up commits", or "annotate the diff" when a large or messy change needs reviewer-ready structure before merge.1.5kinstalls5Review And Shipreview-and-ship is a Cursor plugin skill for systematic pre-deploy review on the current branch. The workflow gathers diff context against the base branch, uncommitted changes, recent commits, and user intent, then runs targeted tests for changed behavior or documents coverage gaps. It reviews for correctness, regressions, security, and intent fit—using parallel subagents on larger diffs—fixes critical issues, commits focused work, and opens or updates a pull request. Developers reach for review-and-ship when a feature branch is ready for human review and merge rather than ad-hoc diff skimming without test verification.1.4kinstalls6What Did I Get Donewhat-did-i-get-done is a Cursor plugin skill that produces short, high-signal summaries of completed coding work for a user-specified window such as yesterday, the last three days, or last week. The workflow resolves the requested range into concrete dates, reads commits authored by the current git user email, excludes merge commits and uncommitted changes, and synthesizes the most important shipped changes into a concise status update that includes the actual date range used. Developers reach for what-did-i-get-done before standups, weekly reports, or handoffs when chat history is incomplete but the git log holds the ground truth.1.4kinstalls7Fix Cifix-ci is a Cursor plugin skill that resolves failing pull request checks by querying GitHub PR status, reading GitHub Actions logs or external check links, and applying the smallest safe fix one failure at a time. The workflow runs gh pr checks --json name,bucket,state,workflow,link, extracts the first actionable error, patches code, pushes, and re-checks until all checks pass. Developers reach for fix-ci when a branch or PR CI is red and they want an iterative path to green without broad refactors. Guardrails enforce one actionable failure per iteration and prefer low-risk changes before larger edits.1.4kinstalls8Pr Review Canvaspr-review-canvas is a Cursor plugin skill that turns pull request reviews into a visual canvas layout so developers can see where changes sit in the codebase instead of scrolling endless unified diffs. The skill is published in cursor/plugins on skills.sh and has 440 installs with a catalog rank of 6. It targets PR review sessions in Cursor where spatial context, file relationships, and change clustering matter more than raw line counts. Developers reach for pr-review-canvas when a PR spans many files, touches related modules, or needs a map-like overview before approving, requesting changes, or writing review comments.1.4kinstalls9Get Pr Commentsget-pr-comments is a Cursor plugin skill that resolves the active pull request for the current branch, pulls GitHub review comments and discussion threads, and returns a concise, severity-grouped summary with a prioritized action list. Instead of switching between the IDE and GitHub in a browser, developers get grouped feedback, ordered next steps, and flagged open questions that still need clarification from reviewers. get-pr-comments fits the moment a feature branch is under review and scattered inline comments, approval notes, and thread replies must be triaged quickly. The workflow targets actionable remediation: critical blockers surface first, nitpicks are separated, and unresolved questions are explicit so merge readiness is clear.1.4kinstalls10Weekly Reviewweekly-review is a Cursor plugin skill that builds executive-readable weekly recaps from git history for status updates, retros, and planning. It reads the current git user email, collects authored commits from the last 7-10 days on the primary branch, excludes merge commits, and groups changes into 2-5 concise bullets plus a classification paragraph covering likely bug fixes, tech debt, and net-new work. Developers reach for weekly-review when they need a factual shipped-work summary without manually scanning a week of commit logs.1.3kinstalls11Check Compiler Errorscheck-compiler-errors is a Cursor plugin skill that automates compile and type-check validation for TypeScript, Rust, and Go repositories. The workflow runs the repo's native build and type-check commands, groups errors by file and category, fixes highest-confidence issues first, and re-runs checks until clean or blocked. Developers reach for check-compiler-errors when local validation or CI is failing on compiler or type errors and they want the agent to surface, summarize, and iteratively clear blockers instead of manually scanning terminal output. Output includes current check status, a grouped error summary, fixes applied, and any remaining blockers.1.3kinstalls12Loop On Ciloop-on-ci is a Cursor plugin skill for developers who need pull request CI to go green without manual retry loops. The skill resolves the PR for the current branch, reads the full PR-attached check set via `gh pr checks` (not just GitHub Actions runs from `gh run list`), diagnoses existing failures first, then watches pending checks with `gh pr checks --watch --fail-fast`. After each fix and push, it re-inspects the entire check set and repeats until every required check passes or a configured limit stops the loop. Reach for loop-on-ci when a branch or PR keeps failing mixed CI providers—required status checks, Actions, and third-party apps—and you want an agent-driven fix-and-verify cycle instead of babysitting the Checks tab.1.3kinstalls13New Branch And Prnew-branch-and-pr is a Cursor plugin skill for a clean branch-and-pull-request workflow. It triggers when starting work that should ship through a dedicated branch and PR. The workflow ensures a clean or explicitly handled working tree, creates a descriptive branch from latest main, completes implementation and tests, commits focused changes, pushes, and opens a concise PR with summary and test notes. Guardrails keep branch scope to one change set and require verification notes before review. Outputs include the branch name, PR summary, test notes, and PR URL. Developers in Cursor reach for this skill to avoid ad hoc branch naming and incomplete PR descriptions.1.3kinstalls14Workflow From Chatsworkflow-from-chats from cursor/plugins cursor-team-kit extracts reusable working preferences from recent Cursor chats instead of summarizing conversations. By default it scans the last seven days of parent transcripts plus relevant subagent evidence, hunting markers like I prefer, always, never, stop, review, PR, CI, and skill while redacting secrets and transcript paths. It clusters preference atoms by workflow shape—shipping, review, debugging, capture, communication—and chooses whether to emit a new skill, rule edit, workflow doc, or no artifact based on confidence ratings strong, medium, weak, or contradicted. Reach for workflow-from-chats when asked to learn preferences, mine feedback, personalize workflows, or generate person-specific or team-specific agent guidance from Cursor history.1.3kinstalls15Run Smoke TestsRun Smoke Tests is a skill from cursor/plugins for end-to-end smoke verification of Cursor plugins using Playwright. The workflow builds prerequisites, runs npm run smoketest or a focused spec such as path/to/test.spec.ts, inspects traces and logs on failure, applies minimal fixes, and reruns until stable. It documents npm run smoketest-no-compile for faster iteration when build artifacts already exist. Developers reach for it after plugin changes to confirm activation, commands, and integrations still work before merge or release.1.3kinstalls16Control Uicontrol-ui is a Cursor plugins skill for building or adapting a local browser or Chrome DevTools Protocol harness to drive and inspect web, IDE, or Electron UIs. It reuses existing Playwright, browser, or Electron harnesses when present; otherwise it assembles a temporary harness around a dev server or Chromium debug port. Developers use control-ui to reproduce UI bugs involving focus, keyboard input, scrolling, and rendering, and to verify visual or accessibility changes with screenshots, a11y snapshots, perf profiles, and visual diffs. Reach for control-ui when chat-only verification is insufficient and browser evidence is required.1.3kinstalls17Control Clicontrol-cli is a Cursor plugin skill for exercising interactive command-line and terminal UIs through a repeatable local harness. It reuses an existing repo test or demo harness when available, otherwise assembles one from standard local tools to reproduce bugs, verify keyboard flows, capture before-and-after transcripts, and check startup regressions, memory leaks, hangs, and resize behavior. Developers reach for control-cli when a CLI or TUI needs systematic validation instead of manual terminal poking, especially for prompt sequences, interrupts, and layout checks before release.1.3kinstalls18Verify Thisverify-this is a verification skill that turns “prove it works” into a falsifiable claim with repeatable evidence. verify-this restates the claim in testable terms, captures a baseline and a treatment run against a running app or service, and compares concrete artifacts (routes, UI behavior, API responses, edge cases, and measurements). verify-this returns one of three verdicts—VERIFIED, NOT VERIFIED, or INCONCLUSIVE—so a developer can decide whether the change is actually correct. Developers reach for verify-this when tests pass but user-visible behavior is uncertain, when a bug fix needs a before/after repro, or when performance or memory claims require measurement instead of a recap.1.2kinstalls19Thermo Nuclear Reviewthermo-nuclear-review is a Cursor plugin skill designed for deep branch audits when standard review is insufficient. The skill sets disable-model-invocation and prompts a security-expert persona to audit the checked-out branch extremely thoroughly for bugs, regressions that break existing features, security vulnerabilities, developer-experience regressions, and feature-gate leaks. Developers invoke it for thermo nuclear, thermonuclear, or deep review requests on large or high-stakes diffs. It stress-tests architecture, edge cases, and failure modes rather than offering lightweight lint feedback, making it appropriate immediately before merging risky PRs.892installs20Thermosthermos is a Cursor plugins skill that runs condensed thermo-nuclear branch review for medium-risk PRs. The workflow scopes changed files, gathers `git diff` context, and launches two background subagents in one message: `thermo-nuclear-review-subagent` for bugs, breakages, security, devex regressions, and feature-flag leaks, plus `thermo-nuclear-code-quality-review-subagent` for maintainability, file-size growth, spaghetti, and abstraction risks including the 1k-line rule. After both finish, thermos deduplicates overlapping findings, prioritizes evidence with file references, and returns a unified verdict. The plugin bundles 3 skills and 2 review subagents migrated from cursor-team-kit. Developers reach for thermos when a branch needs aggressive dual-pass review faster than running the full extended thermo playbook manually.602installs21UnslopCursor plugin skill Unslop cleans agent-generated code slop: it identifies templated noise, over-engineered branches, and comment bloat, then refactors toward concise, idiomatic implementations suitable for production merge.558installs22Typescript Best PracticesTypeScript-best-practices skill encodes Cursor plugin coding standards: strict types, disciplined modules, safe async patterns, and maintainable structures for extension UI, commands, and shared library code.531installs23ArchitectCursor architect plugin helps teams draft clear system designs—components, data flows, constraints, and phased plans—so agents and developers implement features with aligned boundaries, fewer rewrites, and documented tradeoffs.527installs24WhyThe why skill from cursor/plugins prompts deliberate rationale discovery before execution: articulate the problem, expected outcome, and reason a task matters so Cursor agents and developers scope work correctly, avoid solution-first drift, and leave a traceable decision trail for later review.509installs25Interrogateinterrogate is a Cursor plugin skill that runs structured pre-build questioning—goals, constraints, edge cases, and success metrics—so agents and developers lock scope and avoid misimplemented features.504installs26Principle Fix Root CausesEncodes a fix-root-causes discipline for Cursor agents: diagnose the underlying failure mechanism instead of patching visible symptoms, validate that the defect cannot recur, and leave evidence so future changes do not reintroduce the same class of bug.455installs27ArenaCursor Arena plugin skill for comparing LLMs and agent setups in-editor, helping developers pick and tune model defaults for coding, review, and automation tasks during build.443installs28Principle Type System DisciplineCursor plugin skill that codifies type-system discipline for day-to-day coding: explicit types at boundaries, no escape hatches, and patterns that keep agent-generated changes structurally sound in TypeScript-heavy repos.439installs29Principle Foundational ThinkingCursor plugin principle skill that forces foundational analysis: articulate the core problem, list assumptions, derive requirements from first principles, and align language so ideation and later specs share a coherent mental model.438installs30Principle Encode Lessons In Structureprinciple-encode-lessons-in-structure from cursor/plugins shows how to bake team learnings into plugin architecture—folder layouts, templates, and checklists—so Cursor agents automatically apply past fixes and avoid repeating documented mistakes without lengthy prompts.437installs31Continual Learningcontinual-learning from cursor/plugins encodes how to build plugins that learn across sessions: structured feedback capture, persistent memory updates, and incremental rule refinement so Cursor agents improve behavior over time instead of resetting context each chat.435installs32Principle Boundary Disciplineprinciple-boundary-discipline teaches Cursor plugin authors to enforce tight capability boundaries: clear scopes, explicit delegation, and refusal paths so agents do not overreach across tools, reducing unsafe automation and confusing cross-skill behavior.435installs33Principle Experience FirstExperience-first principle for Cursor plugin development: prioritize clear, low-friction developer interactions in commands, panels, and workflows before expanding scope or adding configuration surface area.435installs34Principle Exhaust The Design SpaceCursor plugin principle skill pushing teams to enumerate viable design alternatives, document constraints and tradeoffs, and only then narrow—so validation prototypes test the strongest option rather than anchoring on the initial guess.434installs35Figure It Outfigure-it-out equips the Cursor agent to investigate ambiguous problems via targeted spikes, reproductions, and evidence gathering so teams validate feasibility and narrow scope before committing to full production implementation.431installs36ReflectCursor plugin skill guiding structured reflection after delivery: summarize outcomes versus goals, note surprises and bottlenecks, propose process tweaks, and archive insights so operate-phase learning continuously sharpens future builds.431installs37Principle Laziness ProtocolDefines a laziness protocol for coding agents: maximize reuse of existing code and tooling, postpone nonessential work, and resist redundant implementation so each change stays small, maintainable, and consistent with project conventions.429installs38Principle Guard The Context WindowCursor plugin principle skill teaching context-window discipline: prioritize essential sources, compress stale thread history, externalize bulky artifacts, and monitor token use so agents remain precise through lengthy build sessions.426installs39Principle Build The Leverprinciple-build-the-lever teaches agents to invest in durable leverage—skills, scripts, MCP servers, and workflow automation—instead of repeating one-off manual fixes. It guides prioritizing high-frequency bottlenecks, encoding solutions once, and compounding throughput across builds, reviews, and shipping.423installs40Principle Make Operations IdempotentMake-operations-idempotent principle for Cursor plugins: structure writes, API calls, and state transitions so repeated execution produces the same outcome and never double-applies destructive changes.420installs41Principle Redesign From First PrinciplesPromotes redesign from first principles when incremental fixes accrue debt: restate the real problem, question inherited structure, propose a simpler architecture aligned to requirements, and choose refactor over endless patch stacks when complexity outgrows value.419installs42Principle Prove It WorksProve-it-works principle for Cursor plugin work: no feature is done until verified with executed tests or reproducible demonstrations that show the advertised behavior in a real environment.411installs43Principle Subtract Before You AddSubtract-before-you-add principle for Cursor plugins: remove redundant commands, settings, and code paths before adding features so the extension stays minimal, testable, and easier for users to adopt.410installs44Principle Minimize Reader LoadGuides authors and agents to minimize reader cognitive load: write code and diffs so intent is obvious on first pass, reduce indirection and incidental edits, and structure changes so reviewers can approve confidently without reconstructing hidden reasoning.409installs45Principle Migrate Callers Then Delete Legacy ApisTeaches a safe API deprecation workflow: locate every caller of a legacy interface, move each to the new API, confirm zero remaining references, then delete the old surface so refactors do not break downstream services or clients.404installs46Automate MeCursor automate-me plugin packages repetitive editor and repository workflows into agent-driven automations—templates, scripted steps, and guardrails—so teams repeat complex tasks reliably without re-prompting from scratch each time.403installs47Principle Outcome Oriented ExecutionDirects agents toward outcome-oriented execution: articulate the finished result upfront, choose actions that advance that result, ignore busywork that does not change user-visible value, and declare completion only when concrete acceptance criteria are satisfied.400installs48Show Me Your WorkShow-me-your-work skill instructs Cursor plugin agents to display reasoning, commands run, and artifacts produced so developers can verify automated changes instead of accepting opaque black-box edits.400installs49Principle Never Block On The HumanEstablishes an autonomy rule for Cursor agents: never halt execution solely because a human has not replied—pursue verifiable progress, record assumptions and blockers clearly, and keep momentum while flagging what truly requires human decision or secrets.399installs50Principle Separate Before Serializing Shared StateEncodes a backend concurrency and persistence rule: separate shared state into distinct ownership domains before serialization or transmission, so agents avoid monolithic mutable blobs, narrow failure blast radius, and prevent race conditions across callers and services.396installs51TddGuides test-driven development inside Cursor plugins: start with failing tests, implement minimal code to pass, refactor safely, and repeat. Covers unit and integration patterns so agents ship features with verifiable behavior instead of untested guesses.381installs52Cli For Agentscli-for-agents defines terminal-oriented interfaces for Cursor plugin agents so they can execute scripted workflows, consume structured command output, and chain operations reliably instead of improvising shell usage through prompts alone.372installs53Cursor Sdkcursor-sdk offers a shared integration layer for cursor/plugins authors to access editor APIs, extension hooks, events, and auth patterns without reimplementing low-level wiring in every plugin. It accelerates reliable agent, UI, and automation features across the ecosystem.338installs54OrchestrateCursor plugin skill for orchestrating several agents through a shared plan: decompose work, assign subagents, merge results, and recover from partial failures so complex builds complete with coherent context and minimal manual babysitting.318installs55Create Plugin Scaffoldcreate-plugin-scaffold in cursor/plugins produces starter Cursor plugin projects with correct manifests, skill hooks, command stubs, and folder conventions so developers extend the IDE agent quickly without reverse-engineering layout from existing repos.317installs56Docs Canvasdocs-canvas provides a collaborative documentation canvas for cursor/plugins work so teams and agents can draft, revise, and align READMEs, architecture notes, and runbooks alongside active implementation without losing context in chat alone.315installs57Review Plugin SubmissionReview-plugin-submission skill guides systematic review of Cursor plugin contributions: evaluate diffs, permissions, tests, documentation, and marketplace requirements with consistent criteria before merge or publish.306installs58Ralph LoopCursor plugin skill implementing the Ralph loop pattern for autonomous coding: the agent re-enters work, re-reads context, and continues until acceptance checks pass—aimed at multi-step builds that fail with one-shot prompts.288installs59Create Learning Pathcreate-learning-path helps Cursor users design personalized, milestone-based learning curricula for new languages, frameworks, or roles—linking resources, exercises, and checkpoints before deep implementation work begins.265installs60Poteto Modepoteto-mode is a Cursor plugin interaction preset that compresses agent verbosity and favors fast, action-first edits during routine iteration—useful when you want minimal narration and quicker turnaround on small changes.264installs61Ralph Loop HelpCompanion help skill for Ralph loop in cursor/plugins: onboarding, troubleshooting, and parameter guidance so users can run iterative agent sessions reliably without guessing termination rules or recovery steps.219installs62Cancel Ralphcancel-ralph provides an emergency stop for Ralph-style autonomous agent loops inside cursor/plugins, letting developers halt iterative agent runs before they waste tokens, produce unwanted diffs, or drift from the task. It is agent lifecycle control for Cursor plugin workflows.218installs63Run Learning RetrospectiveCursor plugin skill to run a learning retrospective after completing work: summarize wins and misses, document decisions, and output specific process changes so agent-assisted teams compound quality over sprints.217installs