> ## Documentation Index > Fetch the complete documentation index at: https://docs.prpm.dev/llms.txt > Use this file to discover all available pages before exploring further. # Self-Improving Packages > How AI assistants can automatically discover and install expertise packages from PRPM to enhance their capabilities ## Overview Self-improving packages enable AI assistants (Claude, Cursor, etc.) to **automatically search and install PRPM packages** when they detect they're working on specialized tasks. Instead of relying only on their base training, the AI can dynamically acquire domain-specific expertise from the PRPM registry. ## How It Works ### 1. Task Detection The AI analyzes the user's request for domain-specific keywords: * **Infrastructure**: aws, pulumi, terraform, kubernetes, docker, beanstalk * **Testing**: playwright, jest, cypress, vitest, e2e * **Deployment**: ci/cd, github-actions, gitlab-ci, workflows * **Frameworks**: react, vue, next.js, express, django, fastapi ### 2. Automatic Package Search When keywords are detected, the AI searches the PRPM registry: ```bash theme={null} prpm search "pulumi aws infrastructure" --limit 5 ``` ### 3. Quality Assessment The AI evaluates packages based on confidence levels: **High Confidence (Auto-suggest)** * ✅ Official packages (`@prpm/*`) * ✅ Featured packages * ✅ High downloads (>1,000) * ✅ Verified authors **Medium Confidence (Present options)** * ⚠️ Community packages (\<1,000 downloads) * ⚠️ Multiple similar packages **Low Confidence (Skip)** * ❌ Unverified packages * ❌ Zero downloads ### 4. User Approval The AI presents findings and asks permission: ``` I found a package that can help: 📦 @prpm/pulumi-infrastructure (Official, 3.2K downloads) - Pulumi + AWS best practices - Cost optimization patterns - Production deployment guides Should I install this to enhance my Pulumi knowledge? ``` ### 5. Installation & Application After approval, the package is installed and immediately applied: ```bash theme={null} prpm install @prpm/pulumi-infrastructure ``` The AI then uses the package's expertise to help with the task. ## Real-World Example ### User Request ``` "Help me deploy a Node.js app to AWS Elastic Beanstalk using Pulumi" ``` ### AI Detection & Response ``` Keywords detected: Elastic Beanstalk, Pulumi, deployment Searching PRPM registry... Found relevant packages: 1. @prpm/pulumi-infrastructure (Official, 5.2K downloads) - Pulumi best practices and AWS patterns 2. @prpm/beanstalk-deploy (Official, 2.1K downloads) - Elastic Beanstalk deployment patterns - GitHub Actions integration Should I install these packages to provide expert guidance? ``` ### After Installation The AI now has access to: * Production-tested Pulumi patterns * Beanstalk deployment best practices * Common pitfalls and solutions * CI/CD integration examples ## Installing Self-Improving Packages ### For Claude Code ```bash theme={null} prpm install @prpm/self-improving-skill ``` This skill teaches Claude to: * Detect task-specific keywords * Search the registry proactively * Evaluate package quality * Request permission before installing * Apply package knowledge immediately ### For Cursor ```bash theme={null} prpm install @prpm/self-improve-cursor ``` This rule enables Cursor to: * Automatically trigger searches on infrastructure/testing tasks * Present top packages with download counts * Install packages after user approval * Load expertise for the current task ## Creating Your Own Self-Improving Package You can create packages that teach AI assistants to be self-improving: ```markdown theme={null} # self-improving-python.md ## When to Use Triggers when detecting Python-specific tasks: - web frameworks: flask, django, fastapi - data science: pandas, numpy, scikit-learn - async: asyncio, aiohttp ## Search Strategy When Python framework detected: 1. Extract framework name 2. Search: `prpm search "python {framework} best-practices"` 3. Filter for packages with >500 downloads 4. Present top 3 options to user ## Example User: "Help me build a FastAPI app" Action: Search "python fastapi best-practices" Result: Suggest @community/fastapi-expert (1.2K downloads) ``` Publish it: ```bash theme={null} prpm publish ``` ## Meta-Dogfooding PRPM uses its own self-improving packages for development: * **`@prpm/pulumi-infrastructure`** → PRPM's infrastructure (74% cost savings) * **`@prpm/github-actions-testing`** → PRPM's workflow validation * **`@prpm/postgres-migrations`** → PRPM's database patterns **Benefit**: Users get the same battle-tested expertise that built PRPM itself. ## Privacy & Security * ✅ All searches query the PRPM registry API * ✅ No personal data collected during searches * ✅ Download tracking only on install (anonymous) * ✅ User approval required before installation * ✅ Packages are scanned for quality and safety ## Best Practices ### 1. Be Proactive Search for packages **before** starting complex tasks, not after encountering errors. ### 2. Verify Quality Always check: * Download counts * Official/verified status * Package description relevance ### 3. Ask Permission Never install packages without explicit user approval. ### 4. Apply Immediately Once installed, use the package knowledge right away on the current task. ### 5. Track Helpfulness Note which packages were useful for future reference. ## Advanced Patterns ### Chaining Package Discovery AI can discover packages that help discover more packages: ```bash theme={null} # Install the package discovery expert prpm install @prpm/self-improving-skill # Now the AI can discover domain-specific packages User: "Build a Kubernetes deployment" AI: *searches for kubernetes packages* AI: *finds @prpm/k8s-expert* AI: *suggests installation* ``` ### Context-Aware Installation AI can choose package format based on the editor: ```bash theme={null} # For Claude Code users prpm install @prpm/pulumi-infrastructure --as claude # For Cursor users prpm install @prpm/pulumi-infrastructure --as cursor ``` ### Temporary vs Permanent Some packages are task-specific, others are project-wide: ```bash theme={null} # Temporary: install for one-time task prpm install @prpm/stripe-webhooks # Permanent: add to project prpm install @prpm/typescript-type-safety ``` ## Limitations 1. **Search Quality**: Results depend on package descriptions and tags 2. **Installation Friction**: Requires user approval (by design) 3. **Context Window**: Very large packages may exceed token limits 4. **Network Required**: Registry search needs internet access ## Related Topics * [Creating Meta Packages](/advanced/meta-packages) * [Package Collections](/cli/workflows#working-with-collections) * [AI Prompt Formats](/concepts/formats)