Enum Commands 

pub enum Commands {
    Doc {
        library: String,
        query: String,
        output: Option<PathBuf>,
        limit: Option<usize>,
        no_llm: bool,
        rag: bool,
    },
    Snippet {
        library: String,
        query: Option<String>,
        output: Option<PathBuf>,
        offline: bool,
        save: Option<String>,
        save_all: bool,
        json: bool,
        limit: Option<usize>,
        no_llm: bool,
        rag: bool,
    },
    Search {
        query: String,
        no_llm: bool,
        output: Option<PathBuf>,
        limit: Option<usize>,
        rag: bool,
    },
    Get {
        id: String,
        output: Option<PathBuf>,
    },
    Cache {
        command: CacheCommands,
    },
    Config {

        show: bool,
        api_key: Option<String>,
        cache_dir: Option<PathBuf>,
        auto_cache: Option<String>,
        cache_ttl: Option<u64>,
        max_cache_size: Option<u64>,
        openai_api: Option<String>,
        anthropic_api: Option<String>,
        groq_api: Option<String>,
        openrouter_api: Option<String>,
        huggingface_api: Option<String>,
        custom_endpoint: Option<String>,
        llm_provider: Option<String>,
        llm_model: Option<String>,
        llm_api: Option<String>,
        rag: Option<String>,
        add_official_domain: Option<String>,
        embedding_provider: Option<String>,
        embedding_api_key: Option<String>,
        embedding_model_path: Option<PathBuf>,
        embedding_dimension: Option<usize>,
    
},
    Index {
        path: String,
        id: Option<String>,
        crawl: bool,
        max_depth: Option<u32>,
        max_pages: Option<u32>,
    },
    Sources {
        command: SourceCommands,
    },
    Open {
        id: String,
        output: Option<PathBuf>,
    },
    Update {
        check: bool,
        force: bool,
    },
    Embedding {
        command: EmbeddingCommands,
    },
    Init,
}

Variants

Doc

📚 Browse comprehensive documentation sections and guides

Fields

library: String

Library name (examples: ‘fastapi’, ‘react@18’, ‘django’)

query: String

Topic to search for within documentation (optional - omit for general docs)

output: Option<PathBuf>

Save documentation to file (auto-detects format)

limit: Option<usize>

Limit number of sections shown (default: 10, use 0 for unlimited)

no_llm: bool

Force retrieval-only mode (disable LLM synthesis even if API key configured)

rag: bool

Search locally indexed documents instead of Context7 API

Snippet

🔍 Search code snippets and examples with AI-powered understanding

ENHANCED SEARCH: • Searches official docs (Context7) + your indexed documents (RAG) • Semantic understanding finds relevant content with different wording • Quote prioritization: “useEffect cleanup” gets 10x higher relevance • Optional AI synthesis provides comprehensive answers with citations

SEMANTIC FEATURES: • “memory leaks” finds: “memory cleanup”, “performance issues”, “leak prevention” • “authentication” finds: “auth”, “login”, “security”, “credentials” • Version-specific: react@18, django@4.2

AI SYNTHESIS: • Configure: manx config –llm-api “sk-your-key” • Get answers: manx snippet “react hooks best practices” • Force retrieval: manx snippet react hooks –no-llm

EXAMPLES: manx snippet react “useEffect cleanup” # Semantic search with phrase priority manx snippet “database pooling” –llm-api # Get AI answer with citations
manx snippet fastapi middleware –no-llm # Raw results only manx snippet python “async functions” –rag # Search your indexed code files

Fields

library: String

Library name (examples: ‘fastapi’, ‘react@18’, ‘vue@3’)

query: Option<String>

Search query for specific code snippets

output: Option<PathBuf>

Export results to file (format auto-detected by extension: .md, .json)

offline: bool

Work offline using only cached results (no network requests)

save: Option<String>

Save specific search results by number (e.g., –save 1,3,7)

save_all: bool

Save all search results to file

json: bool

Export in JSON format instead of Markdown (use with –save or –save-all)

limit: Option<usize>

Limit number of results shown (default: 10, use 0 for unlimited)

no_llm: bool

Force retrieval-only mode (disable LLM synthesis even if API key configured)

rag: bool

Search locally indexed documents instead of Context7 API (requires: manx config –rag-enabled)

Search

🔍 Search official documentation across the web

INTELLIGENT WEB SEARCH: • Prioritizes official documentation sites (docs.python.org, reactjs.org, etc.) • Uses semantic embeddings for relevance matching
• Falls back to trusted community sources with clear notification • Optional LLM verification ensures result authenticity

OFFICIAL-FIRST STRATEGY: • Always searches official sources first (10x relevance boost) • Expands to community sources only if insufficient official results • Transparent fallback notifications: “⚠️ Expanded to community sources”

EXAMPLES: manx search “hydra configuration commands” # Auto-detects LLM availability
manx search “react hooks best practices” # Uses LLM if API key configured manx search “python async await” –no-llm # Force embeddings-only mode manx search “authentication” –rag # Search your indexed documents

Fields

query: String

Search query for official documentation

no_llm: bool

Disable LLM verification (use embeddings-only mode even if API key is configured)

output: Option<PathBuf>

Export results to file (format auto-detected by extension: .md, .json)

limit: Option<usize>

Limit number of results shown (default: 8)

rag: bool

Search locally indexed documents instead of web search (requires: manx config –rag-enabled)

Get

📥 Get specific item by ID (doc-3, section-5, etc.)

Fields

id: String

Item ID from previous search or doc command output

output: Option<PathBuf>

Save retrieved item to file

Cache

🗂️ Manage local documentation cache

Fields

command: CacheCommands

Config

⚙️ Configure Manx settings, API keys, and AI integration

Fields

show: bool

Display current configuration settings

api_key: Option<String>

Set Context7 API key (get one at context7.com)

cache_dir: Option<PathBuf>

Set custom cache directory path

auto_cache: Option<String>

Enable/disable automatic caching (values: on, off)

cache_ttl: Option<u64>

Set cache expiration time in hours (default: 24)

max_cache_size: Option<u64>

Set maximum cache size in MB (default: 100)

openai_api: Option<String>

Set OpenAI API key for GPT models

anthropic_api: Option<String>

Set Anthropic API key for Claude models

groq_api: Option<String>

Set Groq API key for fast inference

openrouter_api: Option<String>

Set OpenRouter API key for multi-model access

huggingface_api: Option<String>

Set HuggingFace API key for open-source models

custom_endpoint: Option<String>

Set custom endpoint URL for self-hosted models

llm_provider: Option<String>

Set preferred LLM provider (openai, anthropic, groq, openrouter, huggingface, custom, auto)

llm_model: Option<String>

Set specific model name (overrides provider defaults)

llm_api: Option<String>

Legacy option - Set LLM API key (deprecated, use provider-specific options)

rag: Option<String>

Enable/disable local RAG system (values: on, off)

add_official_domain: Option<String>

Add custom official documentation domain (format: domain.com)

embedding_provider: Option<String>

Set embedding provider for RAG system (hash, onnx:model, ollama:model, openai:model, huggingface:model, custom:url)

embedding_api_key: Option<String>

Set embedding API key for API-based providers

embedding_model_path: Option<PathBuf>

Set embedding model path for local models

embedding_dimension: Option<usize>

Set embedding dimension (default: 384)

Index

📁 Index local documents or web URLs for RAG search

INDEXING SOURCES: • Local files: manx index ~/docs/api.md • Directories: manx index ~/documentation/
• Web URLs: manx index https://docs.rust-lang.org/book/ch01-01-installation.html

SUPPORTED FORMATS: • Documents: .md, .txt, .docx, .pdf (with security validation) • Web content: HTML pages (auto text extraction)

SECURITY FEATURES: • PDF processing disabled by default (configure to enable)
• URL validation (HTTP/HTTPS only) • Content sanitization and size limits

EXAMPLES: manx index ~/my-docs/ # Index directory manx index https://docs.python.org –crawl # Deep crawl documentation site manx index https://fastapi.tiangolo.com –crawl –max-depth 2 # Limited depth crawl manx index api.pdf –alias “API Reference” # Index with custom alias

Fields

path: String

Path to document/directory or URL to index

id: Option<String>

Optional alias for the indexed source

crawl: bool

Enable deep crawling for URLs (follows links to discover more pages)

max_depth: Option<u32>

Maximum crawl depth for deep crawling (default: 3)

max_pages: Option<u32>

Maximum number of pages to crawl (default: no limit)

Sources

📂 Manage indexed document sources

Fields

command: SourceCommands

Open

🔗 Open a specific documentation section by ID

Fields

id: String

Section ID from previous doc command output

output: Option<PathBuf>

Save opened section to file

Update

🔄 Update Manx to the latest version from GitHub

Fields

check: bool

Check for updates without installing

force: bool

Force update even if already on latest version

Embedding

🧠 Manage embedding models and providers for semantic search

EMBEDDING PROVIDERS: • hash: Hash-based embeddings (default, fast, lightweight) • onnx:model: Local ONNX models (requires download) • ollama:model: Ollama API (requires Ollama server) • openai:model: OpenAI embeddings API (requires API key) • huggingface:model: HuggingFace embeddings API (requires API key) • custom:url: Custom endpoint API

EXAMPLES: manx embedding status # Show current provider and models manx embedding set hash # Use hash-based (default) manx embedding set onnx:all-MiniLM-L6-v2 # Use local ONNX model manx embedding set ollama:nomic-embed-text # Use Ollama model manx embedding download all-MiniLM-L6-v2 # Download ONNX model manx embedding test “sample query” # Test current embedding setup

Fields

command: EmbeddingCommands

Init

🚀 Interactive setup wizard for configuring manx

WHAT IT DOES: • Guides you through initial configuration • Sets up Context7 API for official docs • Configures embedding models for semantic search • Optional LLM provider setup for AI synthesis

PERFECT FOR: • First-time users • Reconfiguring after updates • Testing different providers

EXAMPLE: manx init # Start interactive setup wizard

Trait Implementations

impl FromArgMatches for Commands

fn from_arg_matches(__clap_arg_matches: &ArgMatches) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn from_arg_matches_mut( __clap_arg_matches: &mut ArgMatches, ) -> Result<Self, Error>

Instantiate Self from ArgMatches, parsing the arguments as needed.

fn update_from_arg_matches( &mut self, __clap_arg_matches: &ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.

fn update_from_arg_matches_mut<'b>( &mut self, __clap_arg_matches: &mut ArgMatches, ) -> Result<(), Error>

Assign values from ArgMatches to self.

impl Subcommand for Commands

fn augment_subcommands<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate Self via FromArgMatches::from_arg_matches_mut

fn augment_subcommands_for_update<'b>(__clap_app: Command) -> Command

Append to Command so it can instantiate self via FromArgMatches::update_from_arg_matches_mut

fn has_subcommand(__clap_name: &str) -> bool

Test whether Self can parse a specific subcommand