allenk/gwt-integrations
Developer tools이 페이지는 AI로 자동 번역되었습니다. 기술적인 세부 사항은 원문을 확인하세요. 원문 보기
Claude Code Skill and MCP Server for GeminiWatermarkTool — let AI agents remove Gemini watermarks automatically
개요
AI agent integrations for GeminiWatermarkTool — Claude-like skill, Codex-like skill, and a shared MCP server 🆕 — paired with GeminiWatermarkTool v0.3.1. Default tools try the current Gemini-3.5+ profile and automatically retry with the legacy (pre-3.5) profile on skip, so one call handles mixed-vintage inputs. Two *_legacy MCP tools remain for explicit opt-in. See Scenario 7 below. This repo provides specialized skill entrypoints for different agent families plus one shared MCP server. The skill entrypoints are intentionally specialized. The mcp/ directory remains shared across all clients. This repo uses distinct skill entrypoints: - Root skill layout: for Claude-like agents - skills/gwt/: for Codex-like agents This avoids forcing both agent families to share the exact same prompt wording and installer defaults. In other words, the skill layer is split by agent family, while the MCP layer stays shared.
README
원문 README의 일부입니다. 원문 보기.
gwt-integrations
AI agent integrations for GeminiWatermarkTool — Claude-like skill, Codex-like skill, and a shared MCP server
🆕 Gemini 3.5 watermark profile + auto-fallback — paired with GeminiWatermarkTool v0.3.1. Default tools try the current Gemini-3.5+ profile and automatically retry with the legacy (pre-3.5) profile on skip, so one call handles mixed-vintage inputs. Two *_legacy MCP tools remain for explicit opt-in. See Scenario 7 below.
This repo provides specialized skill entrypoints for different agent families plus one shared MCP server.
| Integration | Works with | How agent invokes GWT |
|---|---|---|
| Root Skill | Claude Code and Claude-like agents | Reads the root SKILL.md → builds CLI args → calls binary directly |
| Codex Skill | Codex and Codex-like agents | Installs from skills/gwt/ → reads that SKILL.md → calls binary directly |
| MCP Server | Claude Code, Cursor, Zed, OpenAI Codex, any MCP client | Calls MCP tools → server.py → binary |
The skill entrypoints are intentionally specialized. The mcp/ directory remains shared across all clients.
Layout Strategy
This repo uses distinct skill entrypoints:
- Root skill layout: for Claude-like agents
skills/gwt/: for Codex-like agents
This avoids forcing both agent families to share the exact same prompt wording and installer defaults.
Install Destination Rules
| Integration path | Situation | Install location |
|---|---|---|
| Codex Skill | Normal use | ~/.codex/skills/gwt/bin/ |
| Claude-like Skill | Normal use | ~/.claude/skills/gwt/bin/ |
| MCP Server | Normal use | Shared server, independent of skill layout |
| MCP Server | Local validation with GWT_LOCAL_VALIDATION=1 |
./bin/ at the repo root |
| Any path | Manual override with python install.py --dir |
The directory you specify |
In other words, the skill layer is split by agent family, while the MCP layer stays shared.
추천 도구
다른 키워드를 입력하거나 필터를 제거해 보세요.
설치
npx skillfish add allenk/gwt-integrations