h3nryprod01

Claude Glm 2tier — AI skill for Claude Code

AI community

Two-tier Claude Code: Opus (Max) plans+reviews, GLM-5.2 (Z.ai flat-fee) codes — keep working past the weekly cap.

How to install Claude Glm 2tier

This entry records only its repository, not the path inside it, so there is no exact command to give. Open h3nryprod01/claude-glm-2tier and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Claude Glm 2tier does

Two-tier Claude Code: Opus (Max) plans+reviews, GLM-5.2 (Z.ai flat-fee) codes — keep working past the weekly cap.

Alternatives in AI

UI UX Pro Max Skill — An AI SKILL that provide design intelligence for building professional UI/UX multiple platforms 45.8k ★
Paritok 4b V1 — Non-destructive compression gateway for AI coding agents 1.4k ★
Fusion Plan — OMC interview (interactive) → 3-round Opus 4.8 + GPT-5.5 fusion deepening → concise .omc/plans/ output → revie 468 ★

README

I create this because I get weekly rate limit on Max x20 just after 3 working 5-hour windows. Disappointed @Anthropic.

claude-glm-2tier

**Keep coding after your Claude weekly cap runs out — without paying a premium model to type.**

(The repo started as a two-tier Opus + GLM split and kept its name. It is now a four-tier setup.)

A tiny, no-dependency setup that gives each model the job it is cheapest at:

Tier	Model	Runs as	Job
1. Main	Claude Sonnet 5.5	your normal `claude` session	talks to you, plans, integrates, reports
2. Worker	MiMo-V2.6-Pro (default), or GLM / Kimi	`claude-mimo` / `claude-glm` / `claude-kimi`	writes code, runs the tests, never says "done"
3. Advisor	Claude Opus 5.5	read-only subagent, fresh every review	independent review of the worker's diff, runs the tests itself
4. Escalation	Claude Fable 5.1	read-only subagent, rare	answers the one question the advisor could not settle

They share one git repo and a plain handoff file (`.claude/memory/activeContext.md`). The worker leaves the work as `READY_FOR_OPUS_REVIEW`; Main launches the advisor on the diff; only if the advisor ends with `ESCALATE: ` does Fable get asked. Nothing reaches you unreviewed.

 Main (Sonnet) ──plan──► Worker (MiMo/GLM/Kimi) ──READY_FOR_OPUS_REVIEW──► Main
                                                                             │ diff + criteria
                                                                             ▼
                                                              Advisor (Opus, read-only)
                                                                             │ only if stuck: ESCALATE:
                                                                             ▼
                                                              Escalation (Fable, read-only)