Agent
Cost-saving delegate backed by the GLM model (Zhipu/Z.ai). It does real repo work — read, write, edit, run commands — but ALWAYS through GLM (via mcpglmglmagent), so the work runs on GLM tokens (10x cheaper than Opus), NOT Claude. Use PROACTIVELY for cheap, well-specified, self-contained subtasks: frontend/UI…
Hook
Runs when you submit a prompt, before the agent sees it, executing glm_router_hook.mjs via node. From djerok/glm-mcp.
Skill Claude CodeCodex
Delegate well-specified, non-sensitive coding, testing, documentation, or text tasks to the GLM MCP server. Do not use for secrets, security work, images, architecture, parallel work, huge context, or long dependent loops.
MCP server Claude CodeCodexCursor +2
GLM (Zhipu/Z.ai) as a 10x cheaper file-editing subagent for AI coding agents. Runs locally from the glm-mcp npm package. Needs 1 environment variable to run.