Release Notes — v1.0.2

This release improves CLI usability, service control, and OpenClaw integration metadata.

Added

• clawvisual help command with:
  • global command overview
  • command-specific help (for convert, status, revise, regenerate-cover, initialize)
• Local service management commands:
  • clawvisual stop
  • clawvisual restart
• CLI config support for Gemini key:
  • CLAWVISUAL_GEMINI_API_KEY (alias: GEMINI_API_KEY)

Improved

• Skill metadata (skills/clawvisual/SKILL.md) aligned for OpenClaw scanning:
  • declared required env vars
  • declared local config/service files
  • declared runtime behavior (local MCP service + external network targets)
• Better fallback clarity:
  • when GEMINI_API_KEY is missing, image generation routes directly to OpenRouter (with clearer logs)
• README updates:
  • added usage hint for clawvisual help
  • noted that image generation may be slower when falling back to OpenRouter without GEMINI_API_KEY
  • added platform examples (X/Twitter, Instagram/INS, Xiaohongshu)

Notes

• 1.0.2 tag now points to the latest docs-aligned commit for this release line.
• If you consume this as an OpenClaw skill, use the updated skills/clawvisual/SKILL.md metadata and command references.

Release Notes — v1.0.1

This release focuses on making clawvisual easy to install and use as a global CLI package.

Highlights

• Added global npm CLI support:
  • Install once: npm install -g clawvisual
  • Run directly: clawvisual <command>
• Added CLI-first setup flow:
  • clawvisual initialize can auto-start local service (when using localhost MCP URL)
  • Prints local Web URL after startup
• Added local CLI config commands:
  • clawvisual set/get/unset/config
  • Supports keys like CLAWVISUAL_LLM_API_KEY, CLAWVISUAL_LLM_MODEL, CLAWVISUAL_MCP_URL
• Improved service detection:
  • clawvisual status now verifies service identity to avoid false positives from other local MCP services
• Unified naming:
  • MCP skill namespace/path changed from clawvisual-mcp to clawvisual
• Improved output/download naming:
  • Downloaded bundle names no longer depend on generated hook titles

Quick Start

npm install -g clawvisual
clawvisual set CLAWVISUAL_LLM_API_KEY "your_openrouter_key"
clawvisual initialize
clawvisual convert --input "https://example.com/article" --slides auto
clawvisual status --job <job_id>

Notes

• Existing users should update skill/script paths from skills/clawvisual-mcp/... to skills/clawvisual/....
• README and SKILL docs are updated accordingly.

clawvisual 1.0.0

clawvisual 1.0.0 is the first stable open-source release of a URL or long-form text to social carousel pipeline for creators, growth teams, and agent workflow automation.
It supports both URL input and Direct Long-form Text input, and converts content into publish-ready carousel assets with copy + visuals. It can also be invoked through API or MCP.

Highlights

• End-to-end async generation pipeline for social carousel output
• Multi-ratio output support: 4:5, 1:1, 9:16, 16:9
• Rich output payload: post_title, post_caption, hashtags, slide copy, visual_prompt, and image_url
• Revision workflow support:
  • rewrite copy
  • regenerate cover
  • regenerate selected slides
• MCP-compatible interface for agent/tool integration
• OpenClaw skill package included for direct skill integration
• Bilingual docs (English + Chinese): quickstart, MCP guide, and use cases
• Open-source baseline hardening:
  • MIT License
  • CONTRIBUTING guide
  • improved README structure, screenshots, and quickstart GIF

What’s Included

• Web UI for generation, review, revision, and export
• REST API endpoints under /api/v1/*
• MCP endpoint at /api/mcp (initialize, tools/list, tools/call)
• Skill package:
  • skills/clawvisual-mcp/SKILL.md
  • skills/clawvisual-mcp/scripts/clawvisual-mcp-client.mjs

Quick Start

npm install
cp .env.local.template .env.local
npm run dev

Open: http://localhost:3000

Minimum required env:

• LLM_API_KEY

Defaulted automatically:

• LLM_API_URL=https://openrouter.ai/api/v1/chat/completions
• LLM_MODEL=google/gemini-3-flash-preview

OpenClaw / MCP Integration

Use clawvisual as a skill in OpenClaw by installing skills/clawvisual-mcp and setting:

• CLAWVISUAL_MCP_URL
• CLAWVISUAL_API_KEY (if API key validation is enabled)

Local check:

npm run skill:clawvisual -- tools

Documentation

• Quickstart Guide for URL or Long-form Text to Social Carousel
• MCP Integration Guide for clawvisual
• Use Cases for Agent Workflow Automation
• Chinese docs are included and cross-linked with English docs

Notes

• This release focuses on stable generation/revision workflows and MCP-based integration.
• Docker image distribution is planned for a follow-up release.