Binary reference

ai-cmd

AI-powered command-line interface for natural language command execution

Version9.8.4Size67.4MBAuthorWarith Al MaawaliLicenseProprietaryWebsitehttps://www.digi77.com

File Information

Property	Value
Binary Name	ai-cmd
Version	9.8.4
Build Date	Not disclosed
Rust Version	1.82.0
File Size	67.4MB
Author	Warith Al Maawali
License	Proprietary
Category	Kodachi Binary
Description	AI-powered command-line interface for natural language command execution
Git Commit	unknown
Metadata Generated	2026-07-11T18:16:32Z
Binary Timestamp	Unknown
JSON Data	View Raw JSON

SHA256 Checksum

7ab758c10b9ed9cbb07d96e8afb2153518f23e551a1b3ed8287401c10ed429c9

Features

#	Feature
1	Advanced functionality for Kodachi OS

Security Features

Feature	Description
Input Validation	Argument parsing via clap; per-command validation is the consumer's responsibility
Rate Limiting	Not provided by cli-core
Authentication	Not provided by cli-core (see online-auth)
Encryption	Not provided by cli-core

System Requirements

Requirement	Value
OS	Linux (Debian-based)
Privileges	root/sudo for system operations
Dependencies	OpenSSL, libcurl

Global Options

Flag	Description
`-h, --help`	Print help information
`-v, --version`	Print version information
`-n, --info`	Display detailed information
`-e, --examples`	Show usage examples
`--json`	Output in JSON format
`-o, --output-format <FORMAT>`	Force output format (text\|json)
`--json-pretty`	Pretty-print JSON output with indentation
`--json-human`	Enhanced JSON output with improved formatting (like jq)
`--fields <FIELD_LIST>`	Select specific fields to include in output (comma-separated)
`--limit <NUMBER>`	Limit number of results returned
`--offset <NUMBER>`	Skip first N results (for pagination)
`-d, --work-dir <PATH>`	Working directory (defaults to auto-detected base directory)
`--port <PORT>`	Set custom port number (1024-65535)
`--log-level <LEVEL>`	Set log level (error\|warn\|info\|debug)
`--verbose`	Enable verbose output
`--quiet`	Suppress non-essential output
`--no-color`	Disable colored output
`--config <FILE>`	Use custom configuration file
`--timeout <SECS>`	Set operation timeout in seconds (optional; no default applied)
`--retry <COUNT>`	Retry attempts (optional; no default applied)

Commands

`query`

Process a natural language query and execute the matching command

Usage:

ai-cmd query "<natural language query>"

Options:

--threshold: Confidence threshold for execution
--dry-run: Preview without execution
--auto-execute: Auto-execute high confidence matches
--engine: AI engine tier: auto, tfidf, onnx, onnx-classifier, llm, mistral, genai, claude
--model-path: Path to GGUF model for local LLM tier
--stream: Stream response tokens in real-time
--temperature: Sampling temperature 0.0-2.0 (default: 0.7)
--model: Model name for GenAI tier (supports local, OpenAI/Codex, Claude, Gemini, OpenRouter routing)
--tor-proxy: Route cloud providers through Tor proxy
--use: Use a specific GGUF model file from models/ directory
--no-gateway, --skip-gateway: Bypass gateway validation (dry-run only; rejected for live execution)

Examples:

ai-cmd query "check network connectivity"

ai-cmd query "start tor service" --dry-run

ai-cmd query "get my ip address" --json

ai-cmd query "block internet" --threshold 0.8

ai-cmd query "check tor" --engine onnx

ai-cmd query "check tor" --engine mistral

ai-cmd query "check tor" --engine mistral --json

ai-cmd query "check tor status" --no-gateway --dry-run

ai-cmd query "check tor status" --skip-gateway --dry-run

ai-cmd query "check tor status" --engine genai --model gpt-4o-mini --dry-run

ai-cmd query "check tor status" --engine genai --model claude-sonnet-4-5 --tor-proxy --dry-run

`interactive`

Start interactive REPL mode for continuous queries

Usage:

ai-cmd interactive

Options:

--threshold: Confidence threshold for execution
--engine: AI engine tier: auto, tfidf, onnx, llm, claude
--model-path: Path to GGUF model for local LLM tier

Examples:

ai-cmd interactive

ai-cmd interactive --threshold 0.8

ai-cmd interactive --engine onnx

ai-cmd interactive --engine claude --threshold 0.7

ai-cmd interactive --engine llm --model-path ./models/phi.gguf

ai-cmd interactive --engine genai

`feedback`

Submit feedback to improve intent classification

Usage:

ai-cmd feedback "<query>" [OPTIONS]

Options:

--correct-intent: Specify the correct intent ID
--correct-command: Specify the correct command
--comment: Add a comment

Examples:

ai-cmd feedback "check network" --correct-intent network_check

ai-cmd feedback "test internet" --correct-command "health-control net-check"

ai-cmd feedback "my ip" --comment "Should show IP address"

`preview`

Preview intent classification without execution

Usage:

ai-cmd preview "<query>"

Options:

--alternatives: Number of alternative matches to show

Examples:

ai-cmd preview "check network"

ai-cmd preview "stop tor" --alternatives 5

ai-cmd preview "block internet" --json

`voice`

Voice input mode for natural language queries

Usage:

ai-cmd voice [OPTIONS]

Options:

--continuous: Enable continuous listening mode
--timeout: Timeout for voice input in seconds
--provider: STT provider: whisper-cpp, vosk, placeholder, auto
--voice: TTS voice name
--speed: Speech speed (words per minute)
--list-devices: List available audio devices
--check-deps: Check voice engine dependencies

Examples:

ai-cmd voice

ai-cmd voice --continuous

ai-cmd voice --timeout 60

ai-cmd voice --provider whisper-cpp

ai-cmd voice --voice en-us --speed 200

ai-cmd voice --list-devices

ai-cmd voice --check-deps

`suggest`

Get proactive command suggestions based on usage patterns

Usage:

ai-cmd suggest [OPTIONS]

Options:

--limit: Number of suggestions to show
--proactive: Show proactive suggestions

Examples:

ai-cmd suggest

ai-cmd suggest --limit 10

ai-cmd suggest --proactive

`workflow`

Execute workflow profiles via natural language queries

Usage:

ai-cmd workflow [OPTIONS]

`tiers`

List all AI engine tiers and their availability status

Usage:

ai-cmd tiers [--json]

Examples:

ai-cmd tiers

ai-cmd tiers --json

`tools`

List all callable AI tools with parameter schemas

Usage:

ai-cmd tools [--json]

Examples:

ai-cmd tools

ai-cmd tools --json

`providers`

List available GenAI providers and model configuration

Usage:

ai-cmd providers [--json]

Examples:

ai-cmd providers

ai-cmd providers --json

`model-info`

Show loaded AI model details and configuration

Usage:

ai-cmd model-info [--json]

Examples:

ai-cmd model-info

ai-cmd model-info --json

`policy`

Show current AI policy (intent thresholds, tool allowlist, risk mode)

Usage:

ai-cmd policy [--json]

Examples:

ai-cmd policy

ai-cmd policy --json

`export-intents`

Export all intents as JSON catalog for shared libraries

Usage:

ai-cmd export-intents [--output FILE] [--json]

Examples:

ai-cmd export-intents

ai-cmd export-intents --output intent-catalog.json

ai-cmd export-intents --json

Operational Scenarios

Scenario-oriented workflows generated from the binary's built-in -e --json examples.

Scenario 1: Basic Usage

Basic natural language queries (safe mode examples)

Step 1: Check network status using natural language

ai-cmd query "check network connectivity" --dry-run

Expected Output: Dry-run preview of matched command

Step 2: Get IP address query result with JSON output

ai-cmd query "what is my ip address" --dry-run --json

Expected Output: JSON with intent and command preview

Scenario 2: Advanced Features

Advanced query options and preview workflows

Step 1: Preview command execution without running

ai-cmd query "start tor service" --dry-run

Expected Output: Dry-run preview of Tor command

Step 2: Show interactive mode usage

ai-cmd interactive --help

Expected Output: Interactive mode options and examples

Step 3: Preview intent matching with alternatives

ai-cmd preview "block internet" --alternatives 5

Expected Output: Intent analysis with alternative matches

Step 4: Use custom confidence threshold with dry-run

ai-cmd query "test dns leak" --threshold 0.8 --dry-run

Expected Output: Dry-run preview gated by threshold

Scenario 3: Feedback System

Feedback command syntax and usage patterns

Step 1: Show feedback syntax for intent correction

ai-cmd feedback "check network" --correct-intent network_check --help

Expected Output: Feedback command usage help

Step 2: Show feedback syntax for command correction

ai-cmd feedback "check tor" --correct-command "tor-switch tor-status" --help

Expected Output: Feedback command usage help

Step 3: Show feedback syntax for comments

ai-cmd feedback "check tor" --comment "Should use tor-status subcommand" --help

Expected Output: Feedback command usage help

Scenario 4: Voice Input

Voice command usage, diagnostics, and device checks

Step 1: Show voice mode usage and options

ai-cmd voice --help

Expected Output: Voice command help

Step 2: Show continuous voice mode usage

ai-cmd voice --continuous --help

Expected Output: Voice command help for continuous mode

Step 3: Show timeout option usage

ai-cmd voice --timeout 10 --help

Expected Output: Voice command help with timeout option

Step 4: Show provider selection usage

ai-cmd voice --provider vosk --help

Expected Output: Voice command help with provider option

Step 5: List available audio input devices

ai-cmd voice --list-devices

Expected Output: Audio device list

Step 6: Check voice dependencies

ai-cmd voice --check-deps

Expected Output: Voice dependency status report

Scenario 5: AI Suggestions

Get command suggestions based on usage patterns

Step 1: Get recent command suggestions

ai-cmd suggest

Expected Output: List of suggested commands

Step 2: Get proactive suggestions

ai-cmd suggest --proactive

Expected Output: Popular or proactive suggestions

Step 3: Limit number of suggestions

ai-cmd suggest --limit 5

Expected Output: Top 5 suggestions

Scenario 6: Interactive Mode

Interactive REPL usage and configuration options

Step 1: Show interactive mode usage

ai-cmd interactive --help

Expected Output: Interactive mode help

Step 2: Show threshold usage in interactive mode

ai-cmd interactive --threshold 0.8 --help

Expected Output: Interactive help with threshold option

Step 3: Show ONNX engine interactive usage

ai-cmd interactive --engine onnx --help

Expected Output: Interactive help with engine option

Step 4: Show Claude interactive syntax

ai-cmd interactive --engine claude --threshold 0.7 --help

Expected Output: Interactive help with Claude engine options

Step 5: Show local LLM interactive syntax

ai-cmd interactive --engine llm --model-path ./models/phi.gguf --help

Expected Output: Interactive help with model-path option

Scenario 7: AI Engine Tiers

Select specific AI engine tiers for query processing

Step 1: Auto-select best available engine tier

ai-cmd query "check tor" --engine auto --dry-run

Expected Output: Dry-run preview using auto-selected tier

Step 2: Use TF-IDF keyword matching tier

ai-cmd query "check tor" --engine tfidf --dry-run

Expected Output: Dry-run preview using tfidf tier

Step 3: Use ONNX semantic matching tier

ai-cmd query "check tor" --engine onnx --dry-run

Expected Output: Dry-run preview using onnx tier

Step 4: Show Mistral tier syntax and options

ai-cmd query "check tor" --engine mistral --help

Expected Output: Query help with mistral engine

Step 5: Show GenAI tier syntax and options

ai-cmd query "check tor" --engine genai --help

Expected Output: Query help with genai engine

Step 6: Show legacy LLM tier syntax

ai-cmd query "check tor" --engine llm --help

Expected Output: Query help with llm engine

Step 7: Show custom model-path syntax

ai-cmd query "check tor" --engine llm --model-path ./models/phi.gguf --help

Expected Output: Query help with model-path option

Step 8: Show Claude tier syntax and options

ai-cmd query "check tor" --engine claude --help

Expected Output: Query help with claude engine

Step 9: ONNX engine with dry-run preview

ai-cmd query "check tor status" --engine onnx --dry-run

Expected Output: Preview matched command without executing

Step 10: ONNX engine dry-run validation

ai-cmd query "test dns leak" --engine onnx --dry-run

Expected Output: Dry-run DNS leak query preview

Step 11: Show all available AI tiers and status

ai-cmd tiers

Expected Output: Tier list with availability indicators

Scenario 8: Mistral.rs Local LLM

Mistral tier command syntax and diagnostics

Step 1: Show local mistral query syntax

ai-cmd query "is my system secure?" --engine mistral --help

Expected Output: Query help with mistral options

Step 2: Show mistral streaming syntax

ai-cmd query "check all services" --engine mistral --stream --help

Expected Output: Query help including stream option

Step 3: Show mistral temperature option usage

ai-cmd query "run dns leak test" --engine mistral --temperature 0.3 --help

Expected Output: Query help including temperature option

Step 4: Show loaded model details and tier metadata

ai-cmd model-info --json

Expected Output: JSON model status

Scenario 9: Ollama / GenAI Provider

GenAI tier syntax and provider discovery

Step 1: Show genai query syntax

ai-cmd query "analyze my network" --engine genai --help

Expected Output: Query help with genai options

Step 2: Show genai model override syntax

ai-cmd query "security audit" --engine genai --model llama3:8b --help

Expected Output: Query help with model option

Step 3: List available providers and models

ai-cmd providers --json

Expected Output: JSON provider list

Step 4: Show OpenAI/Codex model syntax with tor-proxy flag

ai-cmd query "detailed analysis" --engine genai --model gpt-4o-mini --tor-proxy --help

Expected Output: Query help including tor-proxy option

Step 5: Show Claude model syntax (direct key or OpenRouter fallback)

ai-cmd query "check tor status" --engine genai --model claude-sonnet-4-5 --dry-run

Expected Output: Dry-run preview using Claude-compatible cloud routing

Step 6: Show Gemini model syntax (direct key or OpenRouter fallback)

ai-cmd query "check dns status" --engine genai --model gemini-2.0-flash --dry-run

Expected Output: Dry-run preview using Gemini-compatible cloud routing

Scenario 10: Tool Calling

Tool-calling queries and tool catalog inspection

Step 1: Tool-calling query in safe dry-run mode

ai-cmd query "what's my tor status and dns config?" --dry-run

Expected Output: Dry-run preview for multi-tool query

Step 2: List all callable tools with JSON metadata

ai-cmd tools --json

Expected Output: JSON tool definitions

Step 3: Show tools grouped by domain

ai-cmd tools

Expected Output: Human-readable tool catalog

Step 4: Export the full intent catalog to a JSON file for shared libraries

ai-cmd export-intents --output intent-catalog.json

Expected Output: Catalog written to intent-catalog.json with total intent count

Note

Output path is relative to the current directory; absolute paths and '..' are rejected. Omit --output to print to stdout, or add --json for an envelope summary.

Scenario 11: Security-First Routing

Fast-path and fallback-path routing examples

Step 1: Fast-path style query with dry-run

ai-cmd query "check tor status" --dry-run

Expected Output: Dry-run preview for tor status

Step 2: Show slow-path mistral syntax

ai-cmd query "explain tor routing in detail" --engine mistral --help

Expected Output: Query help for mistral explanation path

Step 3: Configuration change flow with safety preview

ai-cmd query "change dns to cloudflare" --dry-run

Expected Output: Dry-run preview of DNS change

Scenario 12: AI Policy

View and manage AI policy configuration including intent thresholds, tool allowlists, and risk mode.

Step 1: Show current AI policy (thresholds, tools, risk mode)

ai-cmd policy

Expected Output: Policy details with intent thresholds and tool list

Step 2: AI policy as JSON for programmatic use

ai-cmd policy --json

Expected Output: JSON with version, thresholds, allowlist, signature status

Scenario 13: Model Management

Download, list, and select GGUF models for local LLM inference

Step 1: Show syntax for selecting a specific GGUF model file

ai-cmd query "check tor" --use Qwen3-1.7B-Q4_K_M.gguf --help

Expected Output: Query help with --use model selector

Note

Use --use without --help when the model file exists in models/

Step 2: Show details of loaded model

ai-cmd model-info --json

Expected Output: Model path, architecture, and capabilities

Step 3: Show all AI tiers including model availability

ai-cmd tiers --json

Expected Output: JSON with tier details and download status

Scenario 14: Linux System Commands

General-purpose Linux system commands via natural language

Step 1: Disk usage request preview

ai-cmd query "how much disk space" --dry-run

Expected Output: Dry-run preview for disk space command

Step 2: Process listing request preview

ai-cmd query "list running processes" --dry-run

Expected Output: Dry-run preview for process listing command

Step 3: Network port check request preview

ai-cmd query "check if port 80 is open" --dry-run

Expected Output: Dry-run preview for port inspection command

Step 4: Safe preview of package update flow

ai-cmd query "update system packages" --dry-run

Expected Output: Dry-run preview of package update

Note

Remove --dry-run to execute after authentication

Step 5: Network diagnostic request preview

ai-cmd query "ping google" --dry-run

Expected Output: Dry-run preview for ping diagnostic

Step 6: Log viewing with explicit TF-IDF engine

ai-cmd query "show system logs" --engine tfidf --dry-run

Expected Output: Dry-run preview using tfidf tier

Step 7: Service management query with strict confidence threshold

ai-cmd query "restart nginx service" --threshold 0.9 --dry-run

Expected Output: Dry-run preview gated by 0.9 threshold

Step 8: Firewall inspection query preview

ai-cmd query "check firewall rules" --dry-run

Expected Output: Dry-run preview for firewall status command

Scenario 15: Confidence Threshold Reference

The --threshold flag controls minimum confidence required for a match. Range: 0.3 to 1.0. Default: 0.6.

Step 1: Minimum threshold for broad matching

ai-cmd query "check tor" --threshold 0.3 --dry-run

Expected Output: Dry-run preview with loose confidence gating

Step 2: Default threshold balance

ai-cmd query "check tor" --threshold 0.6 --dry-run

Expected Output: Dry-run preview with default confidence gating

Step 3: Higher threshold for sensitive operations

ai-cmd query "block internet" --threshold 0.8 --dry-run

Expected Output: Dry-run preview with strict confidence gating

Step 4: Very strict threshold for near-exact matching

ai-cmd query "check tor" --threshold 0.95 --dry-run

Expected Output: Dry-run preview requiring near-exact match

Scenario 16: Workflow Profiles

Execute multi-step workflow profiles for complete operations

Step 1: List all available workflow profiles

ai-cmd workflow --list

Expected Output: Profiles grouped by category (privacy, emergency, network, etc.)

Step 2: Preview privacy workflow profile

ai-cmd workflow "maximum privacy setup" --preview

Expected Output: Workflow details, steps, and services used

Note

Use without --preview to execute after review

Step 3: Preview emergency recovery workflow

ai-cmd workflow "emergency recovery" --preview

Expected Output: Workflow details, steps, and services used

Step 4: Dry-run workflow execution

ai-cmd workflow "setup network" --dry-run

Expected Output: Shows what would be executed without running commands

Step 5: Preview with custom confidence threshold

ai-cmd workflow "tor setup" --threshold 0.7 --preview

Expected Output: Workflow preview if confidence >= 0.7

Step 6: Show workflow registry statistics

ai-cmd workflow --stats

Expected Output: Profile counts, categories, and service usage

Step 7: Preview workflow with custom parameters

ai-cmd workflow "complete anonymity" --param country=us --preview

Expected Output: Workflow with custom parameter values

Note

Pass key=value pairs for profile parameters

Scenario 17: Setup & Model Management

Download AI models and manage engine tiers (via ai-trainer)

Step 1: List all AI engine tiers and their availability

ai-cmd tiers

Expected Output: Tier list showing which engines are available

Note

Shows setup hints for unavailable tiers

Step 2: Show active model paths and availability across AI tiers

ai-cmd model-info

Expected Output: Model inventory for embeddings/LLM backends

Note

Useful before running complex queries

Step 3: Get model inventory in JSON format

ai-cmd model-info --json

Expected Output: Structured model metadata

Note

Suitable for automation checks

Step 4: List available provider backends

ai-cmd providers

Expected Output: Provider list and status

Step 5: List providers in JSON format

ai-cmd providers --json

Expected Output: Structured provider status details

Note

Includes fields useful for health checks

Step 6: Show AI tier readiness in JSON format

ai-cmd tiers --json

Expected Output: Tier availability and setup hints as JSON

Scenario 18: Gateway Core Integration

ai-cmd and ai-gateway share the same registry/policy/executor core for consistent behavior

Step 1: Service-style queries now validate through shared gateway core

ai-cmd query "check tor status" --dry-run --json

Expected Output: Dry-run preview with consistent policy decision

Step 2: Bypass gateway validation and execute directly (primary flag)

ai-cmd query "check tor status" --no-gateway --dry-run

Expected Output: Dry-run preview without gateway validation step

Step 3: Same bypass behavior using alias flag

ai-cmd query "check tor status" --skip-gateway --dry-run

Expected Output: Dry-run preview without gateway validation step

Note

`--skip-gateway` is an alias of `--no-gateway`

Step 4: Skip gateway for commands with long execution times

ai-cmd query "test dns leak" --skip-gateway --dry-run

Expected Output: Dry-run preview bypassing gateway timeout

Step 5: Display policy guard configuration used by shared gateway core

ai-cmd policy --json

Expected Output: JSON policy summary

Note

Confirms deny/allow behavior used during execution

Step 6: List tool metadata exposed through the shared core

ai-cmd tools --json

Expected Output: JSON tool catalog with safety tiers

Note

Use this to verify command availability before query execution

Environment Variables

Variable	Description	Default	Values
`RUST_LOG`	Set logging level	info	error\|warn\|info\|debug\|trace
`NO_COLOR`	Disable all colored output when set	unset	1\|true\|yes (any value disables color)

Exit Codes

Code	Description
3	Permission denied
2	Invalid arguments
0	Success
1	General error
4	Network error
5	File not found