Prompts

Guided workflows for common tasks

Prompts (Guided Workflows)

Pre-configured multi-step workflows that combine resources, tools, and AI reasoning to accomplish complex tasks efficiently.

What are Prompts?

Prompts are intelligent workflows that:

• Guide you through multi-step processes
• Combine multiple tools automatically
• Provide context-aware recommendations
• Leverage AI reasoning for better results

Unlike individual tools, prompts orchestrate multiple operations and make intelligent decisions based on intermediate results.

Available Prompts

troubleshoot_issue

Diagnose and resolve system errors using Arch Wiki knowledge.

Purpose: Help users identify and fix system problems by leveraging Wiki documentation and AI analysis.

Workflow:

1. Extract keywords from error description
2. Search Arch Wiki for relevant documentation
3. Analyze error context
4. Provide context-aware suggestions
5. Link to detailed Wiki sections

When to use:

• System errors or failures
• Service/daemon issues
• Boot problems
• Configuration errors
• Package conflicts

Example Input:

"Help! My computer won't start up, it's just showing a grub rescue prompt and I don't know what to do"

Example Output:

• Extracted keywords: “boot”, “grub”, “rescue”
• Wiki pages: GRUB, Boot process, System rescue
• Diagnostic steps
• Recovery commands
• Prevention tips

What it does:

1. Parses your error description
2. Identifies key technical terms
3. Searches multiple Wiki pages
4. Prioritizes most relevant sections
5. Provides step-by-step solutions
6. Links to additional resources

audit_aur_package

Perform comprehensive pre-installation security audit for AUR packages.

Purpose: Evaluate package safety before installation by analyzing metadata, PKGBUILD, and trust indicators.

Workflow:

1. Fetch AUR package metadata
2. Analyze package metadata risk
3. Retrieve and analyze PKGBUILD
4. Check for malicious patterns
5. Evaluate maintainer reputation
6. Generate security recommendations

When to use:

• Before installing any AUR package
• Investigating suspicious packages
• Security audits
• Verifying package safety

Example Input:

"I want to install yay but I'm worried about security, can you check if it's safe?"

Example Output:

• Package info (votes: 2500+, popularity: high)
• Metadata risk: LOW
• PKGBUILD analysis: No suspicious patterns
• Maintainer: Known contributor
• Recommendation: SAFE TO INSTALL

Security Checks:

• ✓ Vote count validation
• ✓ Maintainer verification
• ✓ Package age check
• ✓ Update frequency analysis
• ✓ PKGBUILD pattern scanning
• ✓ Dependency safety

What it provides:

1. Overall risk assessment
2. Detailed security findings
3. Line-by-line PKGBUILD analysis
4. Trust indicators
5. Installation recommendation
6. Alternative suggestions (if risky)

analyze_dependencies

Plan package installation by mapping dependencies and determining optimal install order.

Purpose: Understand package dependencies and create installation plan for complex setups.

Workflow:

1. Query official repositories
2. Search AUR if needed
3. Map dependency tree
4. Check for conflicts
5. Suggest installation order
6. Identify optional vs required deps

When to use:

• Installing packages with many dependencies
• Resolving dependency conflicts
• Planning complex installations
• Understanding package relationships

Example Input:

"I'm trying to set up Docker for my development work, what packages do I need to install?"

Example Output:

• Repository: extra
• Required dependencies: containerd, runc, bridge-utils
• Optional dependencies: docker-compose, docker-buildx
• Conflicts: None
• Install order:
  1. containerd
  2. runc
  3. bridge-utils
  4. docker
  5. docker-compose (optional)

What it analyzes:

1. Direct dependencies
2. Transitive dependencies
3. Conflicts and replaces
4. Optional dependencies
5. Make dependencies (AUR)
6. Check dependencies (AUR)

Installation Planning:

• Groups packages by type
• Orders by dependency chain
• Flags potential conflicts
• Highlights optional features
• Suggests pre-installation steps

Prompt Usage Patterns

Troubleshooting Pattern

User: "I'm trying to start nginx but it keeps saying permission denied, what's going wrong?"
AI (using troubleshoot_issue):
  → Searches Wiki: "nginx", "systemd", "permissions"
  → Finds relevant sections
  → Provides diagnostic commands
  → Suggests fixes with explanations

Security Audit Pattern

User: "I found this package called 'crypto-miner-helper' on AUR, should I trust it?"
AI (using audit_aur_package):
  → Fetches metadata (votes, maintainer, age)
  → Analyzes PKGBUILD for malicious code
  → Checks maintainer reputation
  → Provides risk assessment
  → Recommends action

Dependency Analysis Pattern

User: "I want to set up a development environment with Docker, what all do I need to install?"
AI (using analyze_dependencies):
  → Checks official repos first
  → Maps dependency tree
  → Identifies optional features
  → Suggests install order
  → Warns about conflicts

Combining Prompts

Prompts can be chained for complex workflows:

Example: Safe AUR Installation

1. audit_aur_package → Check if package is safe
2. analyze_dependencies → Plan installation
3. (Manual) install_package_secure → Execute install

Example: System Issue Resolution

1. troubleshoot_issue → Diagnose problem
2. search_archwiki → Get detailed docs
3. (Manual) Apply suggested fix
4. troubleshoot_issue → Verify resolution

Prompt vs Tool vs Resource

Type	Purpose	Interaction	Examples
Resource	Read data	Passive lookup	archwiki://, aur://*/info
Tool	Single operation	Execute function	search_aur, analyze_pkgbuild_safety
Prompt	Multi-step workflow	Guided process	troubleshoot_issue, audit_aur_package

When to use each:

• Resource: Direct data access, you know exact location
• Tool: Single specific operation needed
• Prompt: Complex task requiring multiple steps and AI reasoning

Learn More

Resources

URI-based data access

Tools

Individual executable functions

Examples

Practical usage scenarios

Installation

Setup and configuration