v0.3.1

New Features

• Video image inputs — ai video now accepts --image / -i image inputs from local paths, file:// URLs, http(s):// URLs and data URLs (#57)

Improvements

• Default output filenames — generated text, image and video files now use response IDs when available and random 8-character IDs otherwise, replacing generic output.* filenames (#59)
• Agent setup cleanup — removed outdated opensrc references from local agent configuration (#58)

Contributors

• @ctate

New Features

• Reference image inputs — ai image and ai text now accept repeatable --image references from local paths, file:// URLs, http(s):// URLs and data URLs
• Vision stdin detection — ai text can distinguish piped image bytes from piped text, enabling image prompts from stdin without treating binary data as prompt text

Contributors

• @ctate

Bug Fixes

• Published CLI binary — packages ai as a Node-targeted dist/index.js bundle and points npm's bin/files metadata at dist, fixing global installs that could not execute the TypeScript source (#52)

Improvements

• Node engine requirement — declares and documents the Node.js 20+ runtime requirement for the CLI package (#52)

Contributors

• @ctate

New Features

• Dynamic model discovery — replaced hardcoded fallback model lists with live fetches from the AI Gateway API, so new models are available immediately without a CLI update (#48)
• Model capabilities & metadata — ai models now shows creator, capabilities, and pricing info for each model (#48)
• Language-image model routing — language models tagged with image-generation (e.g. Gemini) are automatically detected and routed through generateText instead of generateImage (#48)
• Stdin image support for language-image models — piping an image via stdin to a language-based image model now correctly passes the image data via the messages API (#48)

Improvements

• --provider renamed to --creator — aligns with the AI Gateway's creator/model-name convention and reserves --provider for future multi-provider support (#48)
• Deduplicated --json output — ai models --json now returns a flat array where each model appears once with all its capabilities (#48)
• Gateway fetch resilience — failed gateway fetches are no longer permanently cached, so retries work correctly; requests time out after 5 seconds (#48)
• Documentation overhaul — filled gaps across README, web docs, and added a SKILL.md for agent integration (#42)
• Favicon & OG images — added favicon and Open Graph image generation to the web app (#40)
• OG card refresh — updated OG card to match the portless design and removed outdated tagline (#41)

Bug Fixes

• Mobile responsive overflow — fixed layout overflow on the landing page for mobile viewports (#47)

Breaking Changes

• ai completions removed — the shell completions command has been removed (#48)
• --provider flag renamed to --creator — update any scripts using --provider to use --creator instead (#48)

Contributors

• @ctate
• @dancer

Improvements

• npm README — replaced symlink with a real file so the README renders correctly on the npm registry (#38)

Contributors

• @ctate

v0.3.0

New Features

• Reference image inputs — ai image and ai text now accept repeatable --image references from local paths, file:// URLs, http(s):// URLs and data URLs
• Vision stdin detection — ai text can distinguish piped image bytes from piped text, enabling image prompts from stdin without treating binary data as prompt text

Contributors

• @ctate

Bug Fixes

• Published CLI binary — packages ai as a Node-targeted dist/index.js bundle and points npm's bin/files metadata at dist, fixing global installs that could not execute the TypeScript source (#52)

Improvements

• Node engine requirement — declares and documents the Node.js 20+ runtime requirement for the CLI package (#52)

Contributors

• @ctate

New Features

• Dynamic model discovery — replaced hardcoded fallback model lists with live fetches from the AI Gateway API, so new models are available immediately without a CLI update (#48)
• Model capabilities & metadata — ai models now shows creator, capabilities, and pricing info for each model (#48)
• Language-image model routing — language models tagged with image-generation (e.g. Gemini) are automatically detected and routed through generateText instead of generateImage (#48)
• Stdin image support for language-image models — piping an image via stdin to a language-based image model now correctly passes the image data via the messages API (#48)

Improvements

• --provider renamed to --creator — aligns with the AI Gateway's creator/model-name convention and reserves --provider for future multi-provider support (#48)
• Deduplicated --json output — ai models --json now returns a flat array where each model appears once with all its capabilities (#48)
• Gateway fetch resilience — failed gateway fetches are no longer permanently cached, so retries work correctly; requests time out after 5 seconds (#48)
• Documentation overhaul — filled gaps across README, web docs, and added a SKILL.md for agent integration (#42)
• Favicon & OG images — added favicon and Open Graph image generation to the web app (#40)
• OG card refresh — updated OG card to match the portless design and removed outdated tagline (#41)

Bug Fixes

• Mobile responsive overflow — fixed layout overflow on the landing page for mobile viewports (#47)

Breaking Changes

• ai completions removed — the shell completions command has been removed (#48)
• --provider flag renamed to --creator — update any scripts using --provider to use --creator instead (#48)

Contributors

• @ctate
• @dancer

Improvements

• npm README — replaced symlink with a real file so the README renders correctly on the npm registry (#38)

Contributors

• @ctate

v0.2.1

Bug Fixes

• Published CLI binary — packages ai as a Node-targeted dist/index.js bundle and points npm's bin/files metadata at dist, fixing global installs that could not execute the TypeScript source (#52)

Improvements

• Node engine requirement — declares and documents the Node.js 20+ runtime requirement for the CLI package (#52)

Contributors

• @ctate

New Features

• Dynamic model discovery — replaced hardcoded fallback model lists with live fetches from the AI Gateway API, so new models are available immediately without a CLI update (#48)
• Model capabilities & metadata — ai models now shows creator, capabilities, and pricing info for each model (#48)
• Language-image model routing — language models tagged with image-generation (e.g. Gemini) are automatically detected and routed through generateText instead of generateImage (#48)
• Stdin image support for language-image models — piping an image via stdin to a language-based image model now correctly passes the image data via the messages API (#48)

Improvements

• --provider renamed to --creator — aligns with the AI Gateway's creator/model-name convention and reserves --provider for future multi-provider support (#48)
• Deduplicated --json output — ai models --json now returns a flat array where each model appears once with all its capabilities (#48)
• Gateway fetch resilience — failed gateway fetches are no longer permanently cached, so retries work correctly; requests time out after 5 seconds (#48)
• Documentation overhaul — filled gaps across README, web docs, and added a SKILL.md for agent integration (#42)
• Favicon & OG images — added favicon and Open Graph image generation to the web app (#40)
• OG card refresh — updated OG card to match the portless design and removed outdated tagline (#41)

Bug Fixes

• Mobile responsive overflow — fixed layout overflow on the landing page for mobile viewports (#47)

Breaking Changes

• ai completions removed — the shell completions command has been removed (#48)
• --provider flag renamed to --creator — update any scripts using --provider to use --creator instead (#48)

Contributors

• @ctate
• @dancer

Improvements

• npm README — replaced symlink with a real file so the README renders correctly on the npm registry (#38)

Contributors

• @ctate

v0.2.0

New Features

• Dynamic model discovery — replaced hardcoded fallback model lists with live fetches from the AI Gateway API, so new models are available immediately without a CLI update (#48)
• Model capabilities & metadata — ai models now shows creator, capabilities, and pricing info for each model (#48)
• Language-image model routing — language models tagged with image-generation (e.g. Gemini) are automatically detected and routed through generateText instead of generateImage (#48)
• Stdin image support for language-image models — piping an image via stdin to a language-based image model now correctly passes the image data via the messages API (#48)

Improvements

• --provider renamed to --creator — aligns with the AI Gateway's creator/model-name convention and reserves --provider for future multi-provider support (#48)
• Deduplicated --json output — ai models --json now returns a flat array where each model appears once with all its capabilities (#48)
• Gateway fetch resilience — failed gateway fetches are no longer permanently cached, so retries work correctly; requests time out after 5 seconds (#48)
• Documentation overhaul — filled gaps across README, web docs, and added a SKILL.md for agent integration (#42)
• Favicon & OG images — added favicon and Open Graph image generation to the web app (#40)
• OG card refresh — updated OG card to match the portless design and removed outdated tagline (#41)

Bug Fixes

• Mobile responsive overflow — fixed layout overflow on the landing page for mobile viewports (#47)

Breaking Changes

• ai completions removed — the shell completions command has been removed (#48)
• --provider flag renamed to --creator — update any scripts using --provider to use --creator instead (#48)

Contributors

• @ctate
• @dancer

Improvements

• npm README — replaced symlink with a real file so the README renders correctly on the npm registry (#38)

Contributors

• @ctate

v0.1.1

Improvements

• npm README — replaced symlink with a real file so the README renders correctly on the npm registry (#38)

Contributors

• @ctate

v0.1.0

0.1.0

• Initial release of the new ai-cli — a lightweight, agent-native CLI for generating text, images, and video
• Commands: ai text, ai image, ai video, ai models, ai completions
• Multi-model comparison via comma-separated -m flag
• Inline terminal preview for images and video (Kitty graphics protocol)
• H.264 keyframe decoding via OpenH264 WASM — no native dependencies required
• Stdin piping support for chaining commands
• Shell completions for bash, zsh, and fish

v0.0.3

What's New

• Skills support - Install skills from GitHub following the Agent Skills standard
• MCP support - Connect to external tools via Model Context Protocol
• Git integration - Full git commands including AI-powered commit messages
• Rules system - AGENTS.md for global and project-level instructions
• Image paste - Ctrl+V to paste images from clipboard
• Web tools - Search, fetch URLs, weather
• Instant tool responses - No delay after silent tool execution
• Word navigation - Option+Left/Right to jump words
• Unit tests - Tests for core utilities

Run npm install -g ai-cli to update.