Bridge external AI CLIs without MCP server for isolated sessions.
npx skills add https://github.com/VCnoC/clink-standalone --skill clink-standalone
Please help me install this Agent Skill. Skill: clink-standalone Source: https://github.com/VCnoC/clink-standalone/tree/main Command: npx skills add https://github.com/VCnoC/clink-standalone --skill clink-standalone
Standalone Clink Standalone provides a direct bridge to external AI CLIs (gemini, codex, claude) without requiring an MCP server, enabling isolated tool usage within conversations.
Copy this skill into your Claude/Gemini/OpenAI Codex skills directory and start the bridge with the clink.py script.
High-intent search queries and answers about installing and using this skill.
You can bridge external AI CLIs directly without an MCP server by using a Python script to enable isolated context and role-based usage across Gemini, Codex, and Claude. This approach supports multi-CLI orchestration and file references for cross-tool automation.
Orchestrating multiple AI CLIs for coding and planning is best handled through a direct bridge that supports cross-CLI orchestration with isolated sessions and per-tool context. This allows role-based prompts and file reference capabilities across different environments.
Yes, you need Python 3.9+ and pydantic to run this cross-tool automation bridge. Additionally, the respective external AI CLIs (gemini, codex, claude) must be installed and authenticated on your system before initiating the bridge.
Yes, you can maintain isolated context when using different AI CLIs by running isolated CLI sessions with per-tool context. This ensures role-based usage and prevents context bleed across Gemini, Codex, and Claude during multi-CLI orchestration.
A limitation of bridging AI CLIs directly without an MCP server is that it requires specific local dependencies, including Python 3.9+, pydantic, and pre-authenticated CLI installations. It operates as a standalone script rather than a persistent server-side integration.
Yes, direct CLI bridging supports file reference capabilities for AI code reviews. It allows you to pass file references directly to external AI CLIs within isolated sessions, enabling role-based prompts for coding, planning, and review tasks.
BumblebiberInstall and configure the hmem MCP server across system-wide or project-local scopes.
paxlabs-incDelegate coding, refactoring, and pull request reviews to an autonomous CLI agent.
joonlabExecute OpenAI Codex or ChatGPT CLI queries within Claude Code.
HsnSaboorBuild and manage MCP servers in Python or TypeScript.
