Cursor Custom Prompts

What it does

When your agent uses it

• Prompt engineering cursor
• Better cursor prompts
• Cursor instructions
• Cursor prompt templates

Example prompts

• “cursor prompts”
• “prompt engineering cursor”
• “better cursor prompts”
• “/cursor-custom-prompts”

Requirements

• Compatibility (from SKILL.md): Designed for Claude Code
• Pre-approved tools (allowed-tools): Read, Write, Edit, Bash(cmd:*)

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

1. 1State the scope, inputs, constraints, acceptance checks, and explicit non-goals.
2. 2Keep prompts free of credentials, customer data, and contradictory instructions; reference reviewed rules rather than duplicating them.
3. 3Evaluate the prompt against a fixture, inspect the diff/output, and version shared prompts through normal review.
4. 4Retire or revise prompts that produce policy, quality, or security regressions.

What it can do on your machine

Read from SKILL.md and the folder at commit cfae287. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves these tools, so the agent can use them without asking each time:

  • Read
  • Write
  • Edit
  • Bash(cmd:*)

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  No scripts in the folder and no shell commands in SKILL.md (its code samples are yaml).

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  Links to these hosts (documentation or services it may open):

  • docs.cursor.com
  • platform.openai.com

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

• Compatibility

  Designed for Claude Code

  From compatibility in the SKILL.md frontmatter.

Context cost

Cursor Custom Prompts loads about 2.3k tokens when it runs, and up to ~3.8k if it reads all its reference files. Until then it costs about 69 tokens; SKILL.md has 441 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~69

When it runs · the whole SKILL.md, loaded when a task matches

~2.3k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~3.8k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from jeremylongshore/tons-of-skills-marketplace at commit cfae287, republished under its MIT licence (© jeremylongshore). 441 words, ~2,253 tokens.

name

cursor-custom-prompts

description

Create effective custom prompts for Cursor AI using project rules, prompt engineering patterns, and reusable templates. Triggers on "cursor prompts", "prompt engineering cursor", "better cursor prompts", "cursor instructions", "cursor prompt templates".

allowed-tools

Read, Write, Edit, Bash(cmd:*)

compatibility

Designed for Claude Code

version

1.19.0

license

MIT

author

Jeremy Longshore <jeremy@intentsolutions.io>

tags

saas, cursor, cursor-custom

1. CONTEXT   → @-mentions pointing to relevant code
2. TASK      → What you want done (specific, actionable)
3. CONSTRAINTS → Rules, patterns, limitations
4. FORMAT    → How the output should look

@src/api/users/route.ts @src/types/user.ts         ← CONTEXT

Create a new API endpoint for updating user profiles. ← TASK

Constraints:                                         ← CONSTRAINTS
- Follow the same pattern as the users route
- Use Zod for input validation
- Return 400 for invalid input, 404 for missing user
- Only allow updating: name, email, avatarUrl

Return the endpoint code and the Zod schema as       ← FORMAT
separate code blocks.

@[existing-similar-feature] @[relevant-types]

Implement [feature name] following the pattern in [reference file].

Requirements:
- [requirement 1]
- [requirement 2]
- [requirement 3]

Constraints:
- Same error handling pattern as [reference]
- Same file structure as [reference]
- Include TypeScript types for all public interfaces

@[buggy-file] @Lint Errors

Bug: [describe the incorrect behavior]
Expected: [describe correct behavior]
Steps to reproduce: [1, 2, 3]

The error message is: [paste error]

Find the root cause and suggest a fix. Do not change
the public API surface.

@[file-to-review]

Review this code for:
1. Logic errors or edge cases
2. Security vulnerabilities (injection, XSS, auth bypass)
3. Performance issues (N+1 queries, unnecessary re-renders)
4. TypeScript type safety (any casts, missing generics)
5. Naming and readability

List issues as: [severity] [line/area] [description] [suggestion]

@[source-file] @[existing-test-file]

Generate tests for [function/class name] covering:
- Happy path with valid inputs
- Edge cases: empty input, null, undefined, max values
- Error cases: invalid input, missing required fields
- Async behavior: success and failure scenarios

Follow the same test structure as [existing-test-file].
Use [vitest/jest/pytest] assertions.

@[file-to-refactor]

Refactor this code to [goal]:
- [specific change 1]
- [specific change 2]

Do NOT change:
- The public API (function signatures, return types)
- The test behavior (existing tests must still pass)
- External imports

# .cursor/rules/code-generation.mdc
---
description: "Standards for AI-generated code"
globs: ""
alwaysApply: true
---
When generating code, always:
1. Add JSDoc comments on all exported functions
2. Include error handling (never let functions throw unhandled)
3. Use named exports (never default exports)
4. Add `import type` for type-only imports
5. Prefer const arrow functions for pure utilities
6. Use discriminated unions over boolean flags

When generating TypeScript:
- Strict mode: no `any`, no `as` casts without justification
- Prefer `unknown` over `any` for unknown types
- Use `satisfies` operator for type narrowing
- Infer types where TypeScript can; annotate where it cannot

# .cursor/rules/test-patterns.mdc
---
description: "Test generation standards"
globs: "**/*.test.ts,**/*.spec.ts"
alwaysApply: false
---
When generating tests:
- Use describe/it blocks with readable descriptions
- Arrange/Act/Assert pattern (AAA)
- One assertion per test (prefer multiple focused tests)
- Mock external dependencies, not internal utilities
- Use factory functions for test data (not inline objects)
- Name test files: {module}.test.ts colocated with source

@src/services/billing.service.ts

I need to add proration logic for subscription upgrades.

Before writing code, first:
1. List the variables involved (current plan, new plan, billing cycle)
2. Show the proration formula with a concrete example
3. Identify edge cases (upgrade on last day, downgrade, free trial)

Then implement based on your analysis.

Convert these function signatures to the Result pattern:

Example input:
  async function getUser(id: string): Promise<User>

Example output:
  async function getUser(id: string): Promise<Result<User, NotFoundError>>

Now convert these:
- async function createOrder(input: CreateOrderInput): Promise<Order>
- async function deleteAccount(userId: string): Promise<void>
- async function sendEmail(to: string, body: string): Promise<boolean>

Create a React form component for user registration.

DO NOT:
- Use class components
- Use any CSS-in-JS library
- Add client-side validation (server validates)
- Use controlled inputs for every field (use react-hook-form)
- Import anything not already in package.json

Turn 1: "Create a basic Express route for GET /api/products"
Turn 2: "Add pagination with page and limit query params"
Turn 3: "Add filtering by category and price range"
Turn 4: "Add sorting by any field with asc/desc direction"
Turn 5: "Add input validation and comprehensive error responses"

© jeremylongshore, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 8 other files (references) in skills/.curated/cursor-custom-prompts of jeremylongshore/tons-of-skills-marketplace.

• SKILL.md
• references/.cursorrules-integration.md
• references/advanced-techniques.md
• references/best-practices.md
• references/domain-specific-prompts.md
• references/errors.md
• references/examples.md
• references/prompt-engineering-fundamentals.md
• references/prompt-templates.md

Open the folder on GitHubat commit cfae287