| title | CLI Reference |
|---|---|
| description | Comprehensive reference for Ummon command-line options |
This document provides a comprehensive reference for all Ummon command-line options and subcommands.
These options apply to all Ummon commands:
Ummon - A code analysis tool for building knowledge graphs
Usage: ummon [OPTIONS] <COMMAND>
Commands:
index Index a codebase to build or update the knowledge graph
query Query the knowledge graph
assist Get AI-assisted recommendations or suggestions
help Display help for a specific command
Options:
-h, --help Print help
-V, --version Print version
The index command builds or updates the knowledge graph from your codebase.
ummon index [OPTIONS] <PATH>
Arguments:
<PATH> Path to the directory to index
Options:
-f, --full Perform a full rebuild of the knowledge graph
--enable-domain-extraction Enable LLM-based domain model extraction
--domain-dir <DOMAIN_DIR> Specify a custom directory for domain extraction [default: src]
--llm-provider <PROVIDER> LLM provider to use [default: openrouter]
--llm-model <MODEL> LLM model to use
-v, --verbose Enable verbose output
-h, --help Print help
# Incremental update of the knowledge graph
ummon index .
# Full rebuild of the knowledge graph
ummon index . --full
# Index with domain model extraction
ummon index . --enable-domain-extraction
# Specify a custom domain directory
ummon index . --enable-domain-extraction --domain-dir models/The query command searches and analyzes your codebase using the knowledge graph.
ummon query [OPTIONS] <QUERY>
Arguments:
<QUERY> Query string (e.g., "select functions where name like 'auth%'")
Options:
-f, --format <FORMAT> Output format: text, json, csv, tree [default: text]
-l, --limit <LIMIT> Maximum number of results to return [default: 20]
--no-llm Skip LLM and only use direct knowledge graph queries
--type-filter <TYPE> Filter results by entity type
--path <PATH> Filter results by file path pattern
--llm-provider <PROVIDER> LLM provider to use [default: openrouter]
--llm-model <MODEL> LLM model to use
-h, --help Print help
# Query using natural language
ummon query "show all authentication functions"
# Query using structured syntax
ummon query "select functions where name like 'auth%'" --no-llm
# Traversal query with relationship
ummon query "functions calling functions where name like 'validate%'" --no-llm
# JSON output format
ummon query "select functions" --format json
# Filter by type
ummon query "find api" --type-filter function
# Filter by path
ummon query "show all entities" --path src/auth
# Limit results
ummon query "select functions" --limit 10The assist command provides AI-assisted recommendations and file suggestions.
ummon assist [OPTIONS] <PROMPT>
Arguments:
<PROMPT> Description of the task or question
Options:
--suggest-files Suggest relevant files for the given prompt
--top <TOP> Number of suggestions to return [default: 5]
--llm-provider <PROVIDER> LLM provider to use [default: openrouter]
--llm-model <MODEL> LLM model to use
-h, --help Print help
# Get assistance with implementing a feature
ummon assist "implement a user registration function"
# Get file suggestions for a proposed change
ummon assist --suggest-files "fix authentication token validation"
# Adjust the number of suggestions
ummon assist --suggest-files "add payment processing" --top 10Ummon uses environment variables for sensitive configuration:
OPENROUTER_API_KEY: API key for LLM services (required for natural language queries, assistance, and domain extraction)
Ummon returns the following exit codes:
0: Success1: General error2: Configuration error3: File system error4: Database error5: LLM service error