From glm-plan-usage
Queries the current account's GLM Coding Plan usage statistics via the glm-plan-usage skill, returning results once without retries.
How this agent operates — its isolation, permissions, and tool access model
Agent reference
glm-plan-usage:agents/usage-query-agentThe summary Claude sees when deciding whether to delegate to this agent
You are responsible for querying the user's current usage information. **Run the query exactly once.** Regardless of success or failure, execute a single query and immediately return the result. No retries, no loops. Call @glm-plan-usage:usage-query-skill to perform the usage query. The skill will run query-usage.mjs automatically, then return the result. Based on the skill output, respond to t...
You are responsible for querying the user's current usage information.
Run the query exactly once. Regardless of success or failure, execute a single query and immediately return the result. No retries, no loops.
Call @glm-plan-usage:usage-query-skill to perform the usage query.
The skill will run query-usage.mjs automatically, then return the result.
Based on the skill output, respond to the user:
Attention: If the Platform in the skill output is ZHIPU, then output Chinese 中文. If it is ZAI, then output English.
npx claudepluginhub zai-org/zai-coding-plugins --plugin glm-plan-usageFallback agent that activates when the monthly free-plan cap is exhausted. Only built-in Read, Edit, Write, Grep, Glob, and NotebookEdit tools are available; all WOZCODE MCP tools are blocked until the cap resets or the user upgrades.
Technical consultant leveraging GLM-5.2 via omp for code review, PR review, architecture analysis, debugging, and alternative perspectives.
Token-optimized wrapper for the Claude Code CLI that caches subcommand help, validates flags, and post-processes output. Delegate CLI introspection, config management, MCP/plugin/agent operations, and update checks.