Skip to content

Commit c8a1ecf

Browse files
cursoragentjabrena
cursoragent
and
jabrena
committed
Add comprehensive guide for creating Cursor AI markdown prompt links
Co-authored-by: bren <bren@juanantonio.info>
1 parent 9acfb05 commit c8a1ecf

1 file changed

Lines changed: 277 additions & 0 deletions

File tree

‎cursor-prompt-link-guide.md‎

Lines changed: 277 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,277 @@
1+
# Creating Markdown Links to Send Prompts to Cursor AI
2+
3+
This guide explains how to create markdown links that can send prompts to Cursor AI, similar to bugbot functionality.
4+
5+
## Overview
6+
7+
While Cursor doesn't have direct URL scheme support for sending prompts like some other tools, there are several approaches to achieve similar functionality:
8+
9+
## Method 1: Using Cursor Rules with Markdown Templates
10+
11+
### 1. Create a Rule Template System
12+
13+
Create a `.cursor/rules/core-rules/prompt-templates-agent.mdc` file:
14+
15+
```markdown
16+
---
17+
description: Handles template-based prompts and quick actions for common development tasks. Activates when users reference specific template keywords or request standardized operations like bug analysis, code review, or documentation generation.
18+
globs:
19+
alwaysApply: false
20+
---
21+
22+
# Prompt Templates Agent
23+
24+
## Template Keywords
25+
26+
When users mention these keywords, apply the corresponding template:
27+
28+
- `@bugbot` - Analyze code for potential bugs and issues
29+
- `@reviewer` - Perform comprehensive code review
30+
- `@documenter` - Generate documentation for code
31+
- `@optimizer` - Suggest performance optimizations
32+
- `@tester` - Generate test cases
33+
34+
## Bug Analysis Template (@bugbot)
35+
36+
When `@bugbot` is mentioned:
37+
38+
1. Analyze the provided code for:
39+
- Logic errors
40+
- Potential runtime exceptions
41+
- Security vulnerabilities
42+
- Performance issues
43+
- Code smell patterns
44+
45+
2. Provide structured output:
46+
- **Issue Type**: Category of the problem
47+
- **Severity**: High/Medium/Low
48+
- **Description**: Clear explanation
49+
- **Suggested Fix**: Concrete solution
50+
- **Prevention**: How to avoid similar issues
51+
52+
## Examples
53+
54+
<example>
55+
User: "@bugbot analyze this function"
56+
Response: Structured bug analysis with categorized findings
57+
</example>
58+
59+
<example type="invalid">
60+
User: "@bugbot analyze this function"
61+
Response: Generic code review without structured bug analysis format
62+
</example>
63+
```
64+
65+
### 2. Create Markdown Documentation with Quick Links
66+
67+
Create a `quick-actions.md` file in your project:
68+
69+
```markdown
70+
# Quick Action Links for Cursor AI
71+
72+
Click these links to quickly access common AI prompts:
73+
74+
## 🐛 Bug Analysis
75+
**Markdown to copy:** `@bugbot analyze the selected code for potential bugs and security issues`
76+
77+
## 📝 Code Review
78+
**Markdown to copy:** `@reviewer perform a comprehensive code review focusing on best practices and maintainability`
79+
80+
## 📚 Documentation
81+
**Markdown to copy:** `@documenter generate comprehensive documentation for the selected code including examples`
82+
83+
## ⚡ Performance Optimization
84+
**Markdown to copy:** `@optimizer analyze and suggest performance improvements for the selected code`
85+
86+
## 🧪 Test Generation
87+
**Markdown to copy:** `@tester generate comprehensive test cases for the selected code including edge cases`
88+
89+
## Usage Instructions
90+
91+
1. Select the code you want to analyze
92+
2. Copy one of the prompts above
93+
3. Open Cursor Chat (Cmd/Ctrl + L)
94+
4. Paste the prompt and press Enter
95+
96+
## Custom Prompts
97+
98+
You can also create custom prompts by combining keywords:
99+
100+
- `@bugbot @reviewer` - Bug analysis + code review
101+
- `@documenter @tester` - Documentation + test generation
102+
- `@optimizer @reviewer` - Performance review
103+
104+
## Advanced Usage
105+
106+
For more complex scenarios, reference specific files:
107+
```
108+
@bugbot analyze @src/utils/auth.ts for security vulnerabilities
109+
```
110+
```
111+
112+
## Method 2: Using Cursor's @ Symbol System
113+
114+
Cursor has built-in @ symbols that can be used in markdown documentation:
115+
116+
```markdown
117+
# Development Workflow Links
118+
119+
## Quick Actions
120+
121+
### Bug Analysis
122+
```
123+
@Cursor Rules @Files
124+
Analyze the current file for potential bugs, focusing on:
125+
- Logic errors and edge cases
126+
- Security vulnerabilities
127+
- Performance bottlenecks
128+
- Code maintainability issues
129+
```
130+
131+
### Code Review
132+
```
133+
@Recent Changes @Lint Errors
134+
Perform a comprehensive code review of recent changes:
135+
- Check for adherence to coding standards
136+
- Identify potential improvements
137+
- Suggest refactoring opportunities
138+
```
139+
140+
### Documentation Generation
141+
```
142+
@Code @Definitions
143+
Generate comprehensive documentation including:
144+
- Function/class descriptions
145+
- Parameter explanations
146+
- Usage examples
147+
- Return value documentation
148+
```
149+
```
150+
151+
## Method 3: Creating Custom Cursor Modes
152+
153+
You can create custom modes in Cursor that act like specialized bots:
154+
155+
### 1. Create a modes.json file
156+
157+
Create `.cursor/modes.json`:
158+
159+
```json
160+
{
161+
"bugbot": {
162+
"name": "Bug Detective",
163+
"description": "Specialized in finding and fixing bugs",
164+
"systemPrompt": "You are a bug detection specialist. When analyzing code, focus on:\n1. Logic errors and edge cases\n2. Security vulnerabilities\n3. Performance issues\n4. Potential runtime exceptions\n\nAlways provide structured output with severity levels and concrete fixes.",
165+
"model": "claude-3.5-sonnet",
166+
"tools": ["terminal", "file_operations"]
167+
},
168+
"reviewer": {
169+
"name": "Code Reviewer",
170+
"description": "Comprehensive code review specialist",
171+
"systemPrompt": "You are a senior code reviewer. Provide thorough reviews focusing on:\n1. Code quality and best practices\n2. Architecture and design patterns\n3. Performance optimization\n4. Security considerations\n5. Maintainability and readability",
172+
"model": "claude-3.5-sonnet",
173+
"tools": ["file_operations"]
174+
}
175+
}
176+
```
177+
178+
### 2. Reference in Markdown
179+
180+
```markdown
181+
# Quick AI Assistants
182+
183+
## 🐛 Bug Detective
184+
Switch to Bug Detective mode and ask: "Analyze this code for bugs"
185+
186+
## 👨‍💻 Code Reviewer
187+
Switch to Code Reviewer mode and ask: "Review this implementation"
188+
189+
## How to Use
190+
1. Open Cursor Chat
191+
2. Select the appropriate mode from the dropdown
192+
3. Paste your code or reference files with @
193+
4. Ask your question
194+
```
195+
196+
## Method 4: Browser Bookmarklets (Advanced)
197+
198+
For a more "link-like" experience, you can create browser bookmarklets:
199+
200+
```javascript
201+
javascript:(function(){
202+
const prompt = "Analyze the selected code for potential bugs and security issues";
203+
navigator.clipboard.writeText(prompt).then(() => {
204+
alert("Prompt copied! Switch to Cursor and paste in chat.");
205+
});
206+
})();
207+
```
208+
209+
Save this as a bookmark with the title "Copy Bugbot Prompt".
210+
211+
## Best Practices
212+
213+
1. **Use Descriptive Keywords**: Make your prompt templates easy to remember
214+
2. **Structure Your Prompts**: Use consistent formatting for better AI responses
215+
3. **Version Control**: Keep your rules and templates in version control
216+
4. **Team Sharing**: Share prompt templates with your team through documentation
217+
5. **Iterate and Improve**: Refine your prompts based on AI responses
218+
219+
## Example Implementation
220+
221+
Here's a complete example of a markdown file with "bugbot-like" functionality:
222+
223+
```markdown
224+
# AI Assistant Quick Actions
225+
226+
## 🚀 Quick Start
227+
1. Select code in Cursor
228+
2. Copy a prompt below
229+
3. Open Chat (Cmd/Ctrl + L)
230+
4. Paste and run
231+
232+
## 🐛 Bug Detection
233+
```
234+
@Cursor Rules
235+
Act as a bug detection specialist. Analyze the selected code and provide:
236+
237+
**BUGS FOUND:**
238+
- Issue type and severity
239+
- Exact location and description
240+
- Recommended fix
241+
- Prevention strategy
242+
243+
Focus on: logic errors, security issues, performance problems, edge cases.
244+
```
245+
246+
## 📋 Code Review
247+
```
248+
@Recent Changes @Lint Errors
249+
Perform a senior-level code review covering:
250+
251+
**REVIEW CHECKLIST:**
252+
- [ ] Code quality and standards
253+
- [ ] Security considerations
254+
- [ ] Performance implications
255+
- [ ] Maintainability
256+
- [ ] Test coverage needs
257+
258+
Provide specific, actionable feedback.
259+
```
260+
261+
## 📖 Documentation
262+
```
263+
@Definitions @Code
264+
Generate comprehensive documentation:
265+
266+
**DOCUMENTATION STRUCTURE:**
267+
- Overview and purpose
268+
- Parameters and return values
269+
- Usage examples
270+
- Edge cases and limitations
271+
- Related functions/classes
272+
273+
Use clear, professional language.
274+
```
275+
```
276+
277+
This approach gives you bugbot-like functionality within Cursor's existing framework, allowing you to create reusable, shareable prompts that can be easily accessed through markdown documentation.

0 commit comments

Comments
 (0)