Deep Stack Research Prompt Generator
Deep stack analysis tool.
Arguments: $ARGUMENTS
Instructions
Generate a comprehensive deep research prompt for the topic "$ARGUMENTS" tailored to the current project's technology stack.
If no topic is provided, ask the user what topic they want to research (e.g., security, performance, maintainability, scalability, testing, deployment, accessibility).
Phase 1: Detect the Technology Stack
Analyze the current project to identify all technologies and their versions. Check these files:
Backend Detection
| File |
Technology |
composer.json |
PHP ecosystem (Laravel, Symfony, etc.) — check require for framework and PHP version |
package.json |
Node.js ecosystem (Express, Fastify, NestJS, etc.) — check for server-side frameworks |
requirements.txt / pyproject.toml / Pipfile |
Python (Django, Flask, FastAPI) |
Gemfile |
Ruby (Rails, Sinatra) |
go.mod |
Go (Gin, Echo, Fiber) |
Cargo.toml |
Rust (Actix, Axum, Rocket) |
pom.xml / build.gradle |
Java (Spring Boot) |
*.csproj / *.sln |
.NET (ASP.NET Core) |
Frontend Detection
Check package.json dependencies for:
- Frameworks: React, Vue, Svelte, Angular, Solid, Preact, Qwik
- Meta-frameworks: Next.js, Nuxt, SvelteKit, Remix, Astro
- Build tools: Vite, Webpack, Parcel, esbuild, Turbopack
- CSS: Tailwind CSS, Bootstrap, styled-components, Sass, PostCSS
Integration/Middleware Layer
Look for glue technologies:
- Inertia.js (
@inertiajs/vue3, @inertiajs/react, etc.)
- Livewire (in
composer.json)
- HTMX (in
package.json or HTML files)
- Turbo/Hotwire (in
package.json or Gemfile)
- Alpine.js (often paired with Livewire/HTMX)
Database Detection
Check these locations:
.env file: DB_CONNECTION, DATABASE_URL
docker-compose.yml: Database service definitions
config/database.php (Laravel), settings.py (Django), etc.
- ORM configs: Prisma (
schema.prisma), Drizzle, TypeORM, Eloquent
Common databases: MySQL, PostgreSQL, SQLite, MongoDB, Redis, Elasticsearch
Infrastructure/Other Tools
- Caching: Redis, Memcached
- Search: Elasticsearch, Meilisearch, Algolia
- Queues: Redis, RabbitMQ, SQS
- Storage: S3, local filesystem
- Auth: Sanctum, Passport, NextAuth, Auth0, Clerk
Version Detection
For each technology found, extract the version:
package.json: Check dependencies and devDependencies for exact versions
composer.json: Check require section
- Lock files (
package-lock.json, composer.lock) have exact versions
.env or Docker configs may specify database versions
Phase 2: Confirm Stack with User
IMPORTANT: First output the detected stack as a text message to the user. Do NOT put the stack details inside the AskUserQuestion tool — the tool has limited space for display.
Step 1: Display the stack as text output
Output a message like this (with actual detected values):
I detected the following technology stack for this project:
**Backend:**
- Rust 1.75
- Tauri 2.0
**Frontend:**
- TypeScript 5.3
- React 18.2
- Vite 5.0
- Tailwind CSS 3.4
**Database:**
- SQLite (via Tauri)
**Desktop:**
- Tauri 2.0 (Rust backend, webview frontend)
**Other:**
- pnpm (package manager)
Step 2: Then ask for confirmation
AFTER displaying the stack details above, use the AskUserQuestion tool with a simple confirmation question:
- Question: "Does this detected stack look correct?"
- Options: "Yes, looks correct" / "Need corrections"
Wait for user confirmation before proceeding. If the user provides corrections, incorporate them into the stack.
Phase 3: Generate the Deep Research Prompt
Once the stack is confirmed, generate a comprehensive research prompt. The prompt should be structured for a deep research tool (like Claude, Perplexity, or similar).
Prompt Template
Generate output in this format (replace placeholders with actual stack details):
START OF GENERATED PROMPT
I'm working on a web application with the following technology stack:
Backend:
- [List each backend technology with version]
Frontend:
- [List each frontend technology with version]
Integration:
- [List integration/middleware if applicable]
Database:
- [List databases and data stores with versions]
Other:
- [List other notable tools]
I need comprehensive research on [TOPIC from $ARGUMENTS] best practices, patterns, and considerations for this specific stack.
Research Scope
1. Individual Technology Analysis
For each technology in my stack, research:
- [TOPIC] best practices specific to this version
- Known issues, vulnerabilities, or limitations related to [TOPIC]
- Configuration options that affect [TOPIC]
- Common mistakes developers make regarding [TOPIC]
- Version-specific considerations (what changed in recent versions)
2. Integration Points
Research [TOPIC] considerations for these technology combinations:
- [Backend framework] + [Frontend framework] (data flow, state management)
- [Backend] + [Database] (query patterns, connection handling)
- [Integration layer] specifics (if applicable)
- [Any other relevant combinations based on the stack]
Focus on issues that arise specifically from these technologies working together, not just individual concerns.
3. Stack-Specific Patterns
Identify [TOPIC] patterns and architectures that are:
- Recommended for this exact stack combination
- Anti-patterns to avoid with this stack
- Trade-offs specific to these technology choices
4. Real-World Considerations
Research:
- Common [TOPIC] issues reported by developers using this stack
- Production lessons learned
- Scaling considerations related to [TOPIC]
- Monitoring and observability for [TOPIC]
Requested Output Format
Please provide your findings organized as:
Executive Summary
- Top 10 most critical [TOPIC] considerations for this stack
- Priority-ranked action items
Per-Technology Guidelines
- Organized by each technology in the stack
- Specific, actionable recommendations
- Code examples where helpful
Integration Guidelines
- [TOPIC] at the boundaries between technologies
- Data flow considerations
- Common pitfalls when technologies interact
Known Issues & Gotchas
- Version-specific bugs or limitations
- Documented vulnerabilities (for security topics)
- Edge cases to watch for
Checklist
- Actionable audit checklist for [TOPIC]
- Can be used to review existing code
Anti-Patterns
- What NOT to do
- Common mistakes with this stack
- Why they're problematic
Resources
- Official documentation links
- Recommended articles, tutorials, talks
- Tools that help with [TOPIC] for this stack
END OF GENERATED PROMPT
Phase 4: Present the Output
Output the generated prompt as plain text that the user can easily copy.
Before the prompt, add:
Here's your deep research prompt for "[TOPIC]" tailored to your stack. Copy this and paste it into your preferred research tool:
After the prompt, add:
Tip: This prompt works well with Claude, ChatGPT, Perplexity, or similar AI research tools. For best results, use a tool that can search the web for current information.
Topic-Specific Additions
Depending on the topic provided in $ARGUMENTS, emphasize different aspects:
If topic is "security":
- Emphasize CVEs, OWASP Top 10, authentication, authorization, input validation
- Include encryption, secrets management, dependency vulnerabilities
- Request exploit examples and mitigation strategies
If topic is "performance":
- Emphasize profiling, caching strategies, database optimization, lazy loading
- Include bundle size, Core Web Vitals, memory management
- Request benchmarking approaches and monitoring tools
If topic is "testing":
- Emphasize unit, integration, e2e testing strategies for the stack
- Include mocking strategies, test data management, CI/CD integration
- Request coverage goals and testing anti-patterns
If topic is "maintainability":
- Emphasize code organization, naming conventions, documentation
- Include refactoring patterns, technical debt management
- Request code review checklists and architecture patterns
If topic is "scalability":
- Emphasize horizontal/vertical scaling, load balancing, caching
- Include database sharding, microservices considerations
- Request capacity planning and bottleneck identification
If topic is "deployment":
- Emphasize CI/CD, containerization, environment management
- Include rollback strategies, zero-downtime deployments
- Request infrastructure as code and monitoring setup
If topic is "accessibility":
- Emphasize WCAG compliance, screen reader support, keyboard navigation
- Include ARIA patterns, color contrast, focus management
- Request testing tools and audit approaches
Notes
- Always include version numbers — they matter for accurate research
- If you can't detect a version, note it as "[version unknown]" and ask the user
- The generated prompt should be self-contained and not require additional context
- Tailor the integration section to the actual technologies detected (don't include generic examples)
1---2name: deepstack3description: "Detects your project's full technology stack, then generates a comprehensive research prompt tailored to a specific topic like security, performance, or testing."4metadata:5 source: https://skillmd.com/skills/skinnyandbald/deepstack6---78# Deep Stack Research Prompt Generator910Deep stack analysis tool.1112Arguments: $ARGUMENTS1314## Instructions1516Generate a comprehensive deep research prompt for the topic **"$ARGUMENTS"** tailored to the current project's technology stack.1718**If no topic is provided**, ask the user what topic they want to research (e.g., security, performance, maintainability, scalability, testing, deployment, accessibility).1920---2122## Phase 1: Detect the Technology Stack2324Analyze the current project to identify all technologies and their versions. Check these files:2526### Backend Detection2728| File | Technology |29|------|------------|30| `composer.json` | PHP ecosystem (Laravel, Symfony, etc.) — check `require` for framework and PHP version |31| `package.json` | Node.js ecosystem (Express, Fastify, NestJS, etc.) — check for server-side frameworks |32| `requirements.txt` / `pyproject.toml` / `Pipfile` | Python (Django, Flask, FastAPI) |33| `Gemfile` | Ruby (Rails, Sinatra) |34| `go.mod` | Go (Gin, Echo, Fiber) |35| `Cargo.toml` | Rust (Actix, Axum, Rocket) |36| `pom.xml` / `build.gradle` | Java (Spring Boot) |37| `*.csproj` / `*.sln` | .NET (ASP.NET Core) |3839### Frontend Detection4041Check `package.json` dependencies for:42- **Frameworks**: React, Vue, Svelte, Angular, Solid, Preact, Qwik43- **Meta-frameworks**: Next.js, Nuxt, SvelteKit, Remix, Astro44- **Build tools**: Vite, Webpack, Parcel, esbuild, Turbopack45- **CSS**: Tailwind CSS, Bootstrap, styled-components, Sass, PostCSS4647### Integration/Middleware Layer4849Look for glue technologies:50- **Inertia.js** (`@inertiajs/vue3`, `@inertiajs/react`, etc.)51- **Livewire** (in `composer.json`)52- **HTMX** (in `package.json` or HTML files)53- **Turbo/Hotwire** (in `package.json` or `Gemfile`)54- **Alpine.js** (often paired with Livewire/HTMX)5556### Database Detection5758Check these locations:59- `.env` file: `DB_CONNECTION`, `DATABASE_URL`60- `docker-compose.yml`: Database service definitions61- `config/database.php` (Laravel), `settings.py` (Django), etc.62- ORM configs: Prisma (`schema.prisma`), Drizzle, TypeORM, Eloquent6364Common databases: MySQL, PostgreSQL, SQLite, MongoDB, Redis, Elasticsearch6566### Infrastructure/Other Tools6768- **Caching**: Redis, Memcached69- **Search**: Elasticsearch, Meilisearch, Algolia70- **Queues**: Redis, RabbitMQ, SQS71- **Storage**: S3, local filesystem72- **Auth**: Sanctum, Passport, NextAuth, Auth0, Clerk7374### Version Detection7576For each technology found, extract the version:77- `package.json`: Check `dependencies` and `devDependencies` for exact versions78- `composer.json`: Check `require` section79- Lock files (`package-lock.json`, `composer.lock`) have exact versions80- `.env` or Docker configs may specify database versions8182---8384## Phase 2: Confirm Stack with User8586**IMPORTANT: First output the detected stack as a text message to the user.** Do NOT put the stack details inside the AskUserQuestion tool — the tool has limited space for display.8788### Step 1: Display the stack as text output8990Output a message like this (with actual detected values):9192```93I detected the following technology stack for this project:9495**Backend:**96- Rust 1.7597- Tauri 2.09899**Frontend:**100- TypeScript 5.3101- React 18.2102- Vite 5.0103- Tailwind CSS 3.4104105**Database:**106- SQLite (via Tauri)107108**Desktop:**109- Tauri 2.0 (Rust backend, webview frontend)110111**Other:**112- pnpm (package manager)113```114115### Step 2: Then ask for confirmation116117AFTER displaying the stack details above, use the AskUserQuestion tool with a simple confirmation question:118119- Question: "Does this detected stack look correct?"120- Options: "Yes, looks correct" / "Need corrections"121122**Wait for user confirmation before proceeding.** If the user provides corrections, incorporate them into the stack.123124---125126## Phase 3: Generate the Deep Research Prompt127128Once the stack is confirmed, generate a comprehensive research prompt. The prompt should be structured for a deep research tool (like Claude, Perplexity, or similar).129130### Prompt Template131132Generate output in this format (replace placeholders with actual stack details):133134---135136**START OF GENERATED PROMPT**137138I'm working on a web application with the following technology stack:139140**Backend:**141- [List each backend technology with version]142143**Frontend:**144- [List each frontend technology with version]145146**Integration:**147- [List integration/middleware if applicable]148149**Database:**150- [List databases and data stores with versions]151152**Other:**153- [List other notable tools]154155I need comprehensive research on **[TOPIC from $ARGUMENTS]** best practices, patterns, and considerations for this specific stack.156157## Research Scope158159### 1. Individual Technology Analysis160161For each technology in my stack, research:162- **[TOPIC]** best practices specific to this version163- Known issues, vulnerabilities, or limitations related to **[TOPIC]**164- Configuration options that affect **[TOPIC]**165- Common mistakes developers make regarding **[TOPIC]**166- Version-specific considerations (what changed in recent versions)167168### 2. Integration Points169170Research **[TOPIC]** considerations for these technology combinations:171- [Backend framework] + [Frontend framework] (data flow, state management)172- [Backend] + [Database] (query patterns, connection handling)173- [Integration layer] specifics (if applicable)174- [Any other relevant combinations based on the stack]175176Focus on issues that arise specifically from these technologies working together, not just individual concerns.177178### 3. Stack-Specific Patterns179180Identify **[TOPIC]** patterns and architectures that are:181- Recommended for this exact stack combination182- Anti-patterns to avoid with this stack183- Trade-offs specific to these technology choices184185### 4. Real-World Considerations186187Research:188- Common **[TOPIC]** issues reported by developers using this stack189- Production lessons learned190- Scaling considerations related to **[TOPIC]**191- Monitoring and observability for **[TOPIC]**192193## Requested Output Format194195Please provide your findings organized as:1961971. **Executive Summary**198 - Top 10 most critical **[TOPIC]** considerations for this stack199 - Priority-ranked action items2002012. **Per-Technology Guidelines**202 - Organized by each technology in the stack203 - Specific, actionable recommendations204 - Code examples where helpful2052063. **Integration Guidelines**207 - **[TOPIC]** at the boundaries between technologies208 - Data flow considerations209 - Common pitfalls when technologies interact2102114. **Known Issues & Gotchas**212 - Version-specific bugs or limitations213 - Documented vulnerabilities (for security topics)214 - Edge cases to watch for2152165. **Checklist**217 - Actionable audit checklist for **[TOPIC]**218 - Can be used to review existing code2192206. **Anti-Patterns**221 - What NOT to do222 - Common mistakes with this stack223 - Why they're problematic2242257. **Resources**226 - Official documentation links227 - Recommended articles, tutorials, talks228 - Tools that help with **[TOPIC]** for this stack229230**END OF GENERATED PROMPT**231232---233234## Phase 4: Present the Output235236Output the generated prompt as plain text that the user can easily copy.237238Before the prompt, add:239240> **Here's your deep research prompt for "[TOPIC]" tailored to your stack. Copy this and paste it into your preferred research tool:**241242After the prompt, add:243244> **Tip:** This prompt works well with Claude, ChatGPT, Perplexity, or similar AI research tools. For best results, use a tool that can search the web for current information.245246---247248## Topic-Specific Additions249250Depending on the topic provided in `$ARGUMENTS`, emphasize different aspects:251252### If topic is "security":253- Emphasize CVEs, OWASP Top 10, authentication, authorization, input validation254- Include encryption, secrets management, dependency vulnerabilities255- Request exploit examples and mitigation strategies256257### If topic is "performance":258- Emphasize profiling, caching strategies, database optimization, lazy loading259- Include bundle size, Core Web Vitals, memory management260- Request benchmarking approaches and monitoring tools261262### If topic is "testing":263- Emphasize unit, integration, e2e testing strategies for the stack264- Include mocking strategies, test data management, CI/CD integration265- Request coverage goals and testing anti-patterns266267### If topic is "maintainability":268- Emphasize code organization, naming conventions, documentation269- Include refactoring patterns, technical debt management270- Request code review checklists and architecture patterns271272### If topic is "scalability":273- Emphasize horizontal/vertical scaling, load balancing, caching274- Include database sharding, microservices considerations275- Request capacity planning and bottleneck identification276277### If topic is "deployment":278- Emphasize CI/CD, containerization, environment management279- Include rollback strategies, zero-downtime deployments280- Request infrastructure as code and monitoring setup281282### If topic is "accessibility":283- Emphasize WCAG compliance, screen reader support, keyboard navigation284- Include ARIA patterns, color contrast, focus management285- Request testing tools and audit approaches286287---288289## Notes290291- Always include version numbers — they matter for accurate research292- If you can't detect a version, note it as "[version unknown]" and ask the user293- The generated prompt should be self-contained and not require additional context294- Tailor the integration section to the actual technologies detected (don't include generic examples)