CLI Commands Reference
Complete reference for all Apex AI Brain CLI commands.
Command Groups
Foundation Commands
/init
Initialize project context by scanning Salesforce project structure.
sfai /init
Parameters:
- Optional:
--path <project_path>- Specify project location
Output: Project context configuration
/status
Display current project and system status.
sfai /status
Shows:
- Project initialization status
- Available agents
- Loaded context
- Statistics
/help
Display command help and documentation.
sfai /help [command]
Examples:
sfai /help # All commands
sfai /help /plan # Plan command help
Development Commands
/plan
Create a structured development plan for a feature request.
sfai /plan "<feature description>"
Returns: JSON plan with task breakdown, dependencies, and agents
Example:
sfai /plan "Add customer portal with approval workflow"
/fix
Fix or refactor existing code.
sfai /fix "<description of issue or preferred fix>"
Agent Commands
/spawn (or /agent-spawn)
Execute a specific agent for a task.
sfai /spawn <agent-name> "<task>"
sfai /agent-spawn <agent-name> "<task>"
Agents:
- See Agents Overview for complete list
Example:
sfai /spawn apex-architect "Design lead service"
/agents
List all available agents.
sfai /agents
Output: Agent list with descriptions
/agent-kill
Stop a running agent.
sfai /agent-kill <agent-id>
/agent-logs
View agent execution logs.
sfai /agent-logs [agent-name] [options]
Options:
--tail N- Show last N lines--filter <keyword>- Filter by keyword--watch- Watch live
Examples:
sfai /agent-logs # All logs
sfai /agent-logs apex-architect # Specific agent
sfai /agent-logs --tail 50 # Last 50 lines
sfai /agent-logs --filter error # Errors only
Model & Cost Commands
/model
Manage AI model selection.
sfai /model <subcommand>
Subcommands:
list- Show available modelsset <model>- Set default modelinfo- Show current settings
Models:
claude- Claude (default)gemini- Google Geminigpt4- OpenAI GPT-4
Examples:
sfai /model list
sfai /model set gemini
/tokens
Check token usage and costs.
sfai /tokens [options]
Options:
--breakdown- By model breakdown--by-agent- By agent breakdown--budget- Budget status--daily- Daily stats--monthly- Monthly stats--export <file>- Export to file
Examples:
sfai /tokens
sfai /tokens --breakdown
sfai /tokens --export report.json
/cost
Analyze AI costs.
sfai /cost [options]
Options:
--by-model- Cost per model--by-agent- Cost per agent--projection- Projected costs--report <type>- Generate report
Examples:
sfai /cost --by-agent
sfai /cost --report monthly
Memory Commands
/memory
Manage shared memory and knowledge base.
sfai /memory <subcommand> [options]
Subcommands:
list- List all memory entriesget <key>- Retrieve valueset <key> <value>- Store valuedelete <key>- Delete entryclear- Clear all memorysearch <query>- Search memoryexport- Export to fileimport <file>- Import from filesnapshot <name>- Create snapshot
Examples:
sfai /memory list
sfai /memory get "design_output"
sfai /memory set "schema" "{...}"
sfai /memory search "account pattern"
sfai /memory export backup.json
Command Syntax
Basic Format
sfai /<command> [options] [arguments]
Options Format
sfai /command --option value
sfai /command --flag
Examples
# No options
sfai /help
# With options
sfai /tokens --breakdown
# With arguments
sfai /spawn apex-architect "Design service"
# Multiple options
sfai /agent-logs --tail 100 --filter error
Global Options
Available Everywhere
--debug # Enable debug logging
--quiet # Suppress output
--json # JSON output format
--no-color # No colored output
Examples
sfai /status --json
sfai /spawn --debug apex-architect "..."
Exit Commands
Leave the CLI
/exit # Normal exit
Ctrl+C # Interrupt
Command Chaining
Run Multiple Commands
# Sequential
sfai /init
sfai /status
sfai /plan "feature"
# With logical operators
sfai /init && sfai /status && sfai /plan "feature"
Tips & Tricks
Use command history
- Arrow up/down to navigate
- Ctrl+R to search history
Abbreviate if possible
- Many commands have shorter forms
/spawn=/agent-spawn
Save useful outputs
- Use
/memory exportfor backup - Use
/tokens --exportfor billing
Chain dependent tasks
- Store outputs in memory
- Reference in next command
Troubleshooting Commands
Verify Installation
sfai /help # Check commands work
sfai /status # Check project loaded
sfai /agents # Check agents ready
Debug Issues
sfai --debug /status # Verbose output
sfai /agent-logs # Check agent operations
sfai /memory list # Verify memory system
Reset/Cleanup
sfai /memory clear --session # Clear session
sfai /memory cleanup # Cleanup old entries