Changelog

Track the evolution of Delta CLI with our detailed changelog

View releases with downloadsView on GitHubUpdates every hour

Changelog

All notable changes to Delta CLI will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[Unreleased]

[v0.4.8-alpha] - 2025-07-03

Added

  • Man page generation system

    • New :man command for generating and managing Unix man pages
    • Generates standard troff format man pages for Delta and its commands
    • Subcommands: generate, preview, install, view, completions
    • Shell completion generation for bash, zsh, and fish shells
    • Makefile targets: make man, make install-man, make completions
    • PowerShell build script support for man page generation
    • Structured command documentation system for consistent help

  • Memory & Learning System (Milestone 4)

    • Core learning mechanisms for command patterns
    • Real-time pattern learning from command execution
    • Command sequence detection and learning
    • Directory-specific and time-based pattern recognition
    • Feedback collection system with implicit and explicit modes
    • Interactive feedback collection with :learn feedback
    • Daily training pipeline with Docker and local support
    • Training data extraction and evaluation tools
    • Integration with existing memory and inference systems
    • New :learning command with comprehensive subcommands
    • New :training command for data management
    • Automatic model deployment after successful training

  • Shell Completion Enhancements

    • Full zsh completion support with subcommand awareness
    • Full fish completion support with proper command chaining
    • Intelligent completion for Delta's internal commands (prefixed with :)
    • Subcommand completion for complex commands like :ai, :memory, :learning
    • Integration with existing help system via :man completions command

  • Enterprise Update Features - Channel Management System (Phase 5)

    • Advanced channel policies (stable, beta, alpha, nightly, custom)
    • Channel-specific update configurations with fine-grained control
    • Enterprise mode with access control and user restrictions
    • Forced channel management for organizations
    • Scheduled channel migrations with user targeting
    • Channel history tracking and statistics
    • New commands: :update channel <name>, :update channels
    • Channel policy management with version constraints
    • Custom update URLs and verification keys for enterprise channels
    • Integration with existing update system

  • Enterprise Update Features - Metrics & Reporting System (Phase 5)

    • Comprehensive metrics collection for all update operations
    • Real-time tracking of downloads, installations, and rollbacks
    • Channel-specific performance metrics and analytics
    • Version adoption tracking with success rates
    • Error analysis with pattern detection and suggestions
    • Performance statistics including download speeds and durations
    • Export capabilities in JSON, CSV, and Prometheus formats
    • New commands: :update metrics, :update metrics report
    • Configurable data retention and privacy controls
    • System resource monitoring during updates

Fixed

  • Ollama "connection restored" message no longer shows on startup
    • Added first-check detection to prevent misleading restoration messages
    • Connection status messages now only show for actual state changes

Changed

  • Planning and milestone documents moved to private repository
    • Added docs/planning/ and docs/milestones/ to .gitignore
    • Removed sensitive OAuth and device activation plans from public repo

[v0.4.6-alpha] - 2025-06-24

Added

  • Natural language command suggestion system

    • New :suggest command (alias :s) for getting command suggestions from plain text descriptions
    • Pattern-based matching for common command patterns (file operations, git, docker, etc.)
    • Project-aware suggestions based on detected project type (Node.js, Go, Python, etc.)
    • AI-powered suggestions when Ollama is enabled
    • History-based suggestions that learn from your command usage
    • Interactive selection interface with safety indicators
    • Command explanation feature with :suggest explain <command>
    • Safety validation with warnings for dangerous commands
    • Context-aware caching for improved performance
    • Examples: :suggest list files, :suggest install dependencies, :suggest create new branch

  • Ollama health monitoring system for AI features

    • Periodic connectivity checks when AI is disabled
    • Automatic notifications when Ollama becomes available
    • Configurable check intervals and notification settings
    • New commands: :ai health, :ai health monitor on/off, :ai health interval <seconds>, :ai health notify on/off
    • Smart backoff strategy for failed connection attempts
    • Integrated with configuration persistence

Added - Command Validation Phase 5: Advanced Features (2025-06-21)

Overview

Completed Phase 5 of the command validation system, adding advanced features including obfuscation detection, custom rule engine with DSL, Git-aware safety checks, and CI/CD pipeline integration.

Features

1. Obfuscation Detection (validation/obfuscation_detector.go)
  • Detects multiple obfuscation techniques:
    • Base64 encoded commands (e.g., echo "cm0gLXJmIC8=" | base64 -d)
    • Hex encoding (e.g., echo -e "\x72\x6d")
    • Unicode escape sequences
    • Variable substitution tricks (e.g., a="r"; b="m"; $a$b -rf /)
    • Character substitution (e.g., rm${IFS}-rf${IFS}/)
    • Command substitution abuse
    • Eval chains and source from URLs
  • Provides deobfuscated command analysis
  • Risk assessment based on obfuscation techniques used
2. Custom Rule Engine with DSL (validation/custom_rule_engine.go)
  • YAML-based DSL for defining custom validation rules
  • Default rules file at ~/.config/delta/validation_rules.yaml
  • Pre-configured with security rules:
    • Force push protection for main/master branches
    • Curl pipe bash detection
    • Password exposure prevention
    • Docker privileged container warnings
    • AWS credential exposure detection
    • Database drop operation protection
  • CLI commands for rule management:
    • :validation rules list - List all custom rules
    • :validation rules add - Add new rule interactively
    • :validation rules edit <name> - Edit existing rule
    • :validation rules delete <name> - Delete a rule
    • :validation rules enable/disable <name> - Toggle rule status
    • :validation rules test <command> - Test command against rules
3. Git-Aware Safety Checks (validation/git_safety.go)
  • Specialized safety checks for Git operations:
    • Force push detection on protected branches
    • Hard reset warnings with uncommitted changes
    • Aggressive clean operation alerts
    • Sensitive file addition detection (.env, .key, .pem files)
    • Rebase warnings on protected branches
  • Automatic Git context detection
  • Protected branch configuration (main, master, production, release)
4. CI/CD Pipeline Integration (validation/cicd_validator.go)
  • Automatic CI/CD platform detection:
    • GitHub Actions
    • GitLab CI
    • CircleCI
    • Jenkins
    • Travis CI
    • Azure Pipelines
  • Platform-specific validations:
    • Secret exposure prevention in commands
    • Environment variable exposure detection
    • Deprecated command warnings (e.g., GitHub Actions ::set-output)
    • CI-specific dangerous pattern detection
  • Enhanced security for automated environments
5. Configuration Enhancements

New configuration keys:

  • custom_rules - Enable custom validation rules (default: true)
  • obfuscation_detection - Enable obfuscation detection (default: true)

Testing

  • Comprehensive integration tests for all Phase 5 features
  • Unit tests for each component:
    • Obfuscation detection patterns
    • Custom rule engine operations
    • Git safety checks
    • CI/CD platform detection

[v0.4.5-alpha] - 2025-06-21

Added - Command Validation Phase 4: Interactive Safety (2025-06-20)

Overview

Implemented comprehensive interactive safety features for command validation, providing smart confirmation prompts, educational content, and safer alternative suggestions for dangerous commands.

Features

1. Interactive Safety Checker (validation/interactive_safety.go)
  • Smart confirmation prompts for risky operations with multiple user options:
    • [y] Yes, proceed with the command
    • [n] No, cancel the command
    • [m] Modify the command
    • [s] Mark as safe for future (proceed without prompts)
    • [?] Show more details
  • Command safety history tracking with decision recording
  • Trusted path detection to bypass prompts in safe directories
2. Risk-Based Educational Content
  • Different educational content for each risk level:
    • Critical: System destruction warnings with severe consequences
    • High: Dangerous operation alerts with security implications
    • Medium: Caution advisories with potential side effects
    • Low: No prompts (safe operations)
  • Shows potential consequences of dangerous commands
  • Provides safer alternatives with practical examples
  • Links to documentation for additional learning
3. Configuration Options

New configuration keys available via :validation config set <key> <value>:

  • enabled - Enable/disable validation (default: true)
  • interactive_safety - Enable interactive prompts (default: true)
  • educational_info - Show educational content (default: true)
  • auto_deny_critical - Auto-deny critical commands (default: true)
  • bypass_trusted_paths - Skip prompts in trusted directories (default: true)
4. Command Execution Integration
  • Commands are automatically validated before execution
  • Both interactive shell and direct command mode (-c flag) are covered
  • Validation respects configuration settings
  • Non-blocking for safe commands (low risk)
5. New Commands
  • :validation stats - View safety decision statistics
  • :validation history - View recent safety decisions
  • :validation config set/get/reset - Manage configuration
6. Trusted Path Detection
  • Automatically bypasses prompts in trusted directories:
    • ~/projects
    • ~/dev
    • ~/code
    • ~/src
  • Configurable to add more trusted paths

Example Interactive Session

When a user tries to run rm -rf /, they see:

∆ delta
Terminal output display.
===============================================
⚠️ CRITICAL RISK: System Destruction Warning
===============================================
This command could permanently damage your system or delete critical data.
⚠️ Potential Consequences:
• Complete system failure requiring reinstallation
• Permanent loss of all data
• Corruption of system files
• Loss of user accounts and configurations
💡 Safer Alternatives:
Use specific paths instead of /
Example: rm -rf /tmp/specific-folder
Why safer: Limits damage to intended targets only
Use trash command
Example: trash /path/to/file
Why safer: Allows recovery if mistakes are made
📚 Learn more: https://wiki.deltacli.com/safety/critical-commands
------------------------------------------------------------
🔴 Command Risk Summary
Command: rm -rf /
Risk Factors:
• CRITICAL: This command will recursively delete your entire system!
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⚠️ Do you want to proceed with this CRITICAL risk command?
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Options:
[y] Yes, proceed with the command
[n] No, cancel the command
[m] Modify the command
[s] Mark as safe for future (proceed without prompts)
[?] Show more details
Your choice [y/n/m/s/?]:

Technical Implementation

  • Files Added:

    • validation/interactive_safety.go - Core interactive safety implementation
    • command_validator.go - Command validation integration
    • test_interactive_safety.sh - Test script for safety features

  • Files Modified:

    • validation/engine.go - Added interactive safety checker integration
    • cli.go - Integrated validation into command execution flow
    • validation_commands.go - Added stats, history, and config commands
    • config_manager.go - Added custom config storage for validation settings
    • Makefile - Added command_validator.go to build sources
    • CLAUDE.md - Updated documentation
    • TASKS.md - Marked Phase 4 as completed

This completes Phase 4 of the Command Validation system, providing a user-friendly and educational safety layer that helps prevent accidental damage while teaching users about command risks.

[0.4.4-alpha] - 2025-06-20

Added

  • Command Validation System - Phase 1-3 implementation
    • Syntax validation engine for multiple shells
    • 17 built-in safety rules with pattern matching
    • Context-aware risk assessment
    • Permission requirement detection
    • Environmental context analysis
    • Risk mitigation suggestions

Fixed

  • Direct command mode (-c flag) now includes training pipeline for AI/ML features
  • Commands execute immediately with async training in background

[0.4.3-alpha] - 2025-06-20

Fixed

  • i18n feature now works outside source directory by downloading from GitHub releases
  • Added SHA256 checksum verification for secure i18n file downloads

[0.4.2-alpha] - 2025-06-17

Added

  • Auto-Update System Phase 4: Advanced Features
    • Update scheduling with cron-like functionality
    • Comprehensive update history tracking
    • Post-update validation with automatic rollback
    • Interactive update UI with user choices
  • AI System Improvements:
    • AI enabled/disabled state now persists between sessions
    • Added :ai enable and :ai disable aliases
  • i18n System Improvements:
    • Automatic installation of language files
    • Built-in English translations as fallback
    • Robust path detection for i18n files
    • Clear startup notice if translations are missing

Fixed

  • AI configuration persistence issues
  • i18n file installation and detection

[0.4.1-alpha] - 2025-06-08

Added

  • Automatic version injection at build time
  • Git integration for version and commit detection
  • Smart development vs release build detection

[0.4.0-alpha] - 2025-06-08

Added

  • Auto-Update System Phase 3: Download, Installation & Rollback
    • Secure update downloads with SHA256 verification
    • Automatic backup before updates
    • Rollback capability on failure
    • Interactive update prompts

[0.3.1-alpha] - 2025-06-07

Added

  • Auto-Update System Phase 2: GitHub Integration
    • Update detection from GitHub releases
    • Version comparison and notifications

[0.3.0-alpha] - 2025-06-06

Added

  • Auto-Update System Phase 1: Foundation Infrastructure
    • Configuration management for updates
    • Version management system
    • CLI commands for update control

[0.2.0-alpha] - 2025-06-06

Added

  • Internationalization (i18n) support for 11 languages
  • ART-2 machine learning system for pattern recognition
  • Memory collection system for command history
  • Jump command for directory navigation
  • Inference system for learning capabilities

[0.1.0-alpha] - 2025-05-15

Added

  • Initial release
  • AI-powered command suggestions with Ollama integration
  • Encrypted command history
  • Shell compatibility (bash, zsh, fish)
  • Internal command system with : prefix
  • Tab completion support