Route engineering tasks through ACG workflows and GLM-5 via z.ai.
npx skills add https://github.com/EndUser123/cc-marketplace --skill ai-oc-glm51
Please help me install this Agent Skill. Skill: ai-oc-glm51 Source: https://github.com/EndUser123/cc-marketplace/tree/main/plugins/cc-skills-ai-cli/skills/ai-oc-glm51 Command: npx skills add https://github.com/EndUser123/cc-marketplace --skill ai-oc-glm51
OpenCode-powered assistant that grounds research and engineering analysis by routing tasks through an ACG-style workflow and GLM-5 model via z.ai.
Provide a task prompt like 'Analyze the auth module and propose improvements' to trigger the ACG workflow and routing.
High-intent search queries and answers about installing and using this skill.
To route engineering analysis tasks through an ACG workflow with GLM-5, provide a task prompt to trigger soft triage into RESEARCH, ENGINEERING, DESIGN, or RCA, which then deterministically routes to the appropriate workflow via z.ai.
OpenCode can perform root-cause investigations and design decisions by grounding analysis in source material and routing prompts through the ACG workflow and GLM-5 model to deliver structured outputs with source-fidelity citations.
The best way to triage codebase tasks into research or engineering paths is using soft triage to categorize prompts into RESEARCH, ENGINEERING, DESIGN, or RCA, with user override options for deterministic routing.
The engineering path supports automated verification through a verification pyramid that includes Tier 1 Unit, Tier 2 Integration, and Tier 3 E2E testing, ensuring source fidelity and structured outputs.
To override the default triage routing for a design analysis task, use the user override feature during the soft triage phase to manually select DESIGN, redirecting the prompt to the Hypothesis Ledger or appropriate workflow.
ACG workflows for codebase analysis enforce strict source fidelity and frontmatter-defined processes, limiting outputs to structured formats and explicit path choices, which may restrict freeform exploratory coding tasks.
dgk-devAutomate GLM-5 code review validation and triage with glm-review.
DatTran26Execute unit, integration, e2e, and UI tests with coverage reports.
twikusAutomate software delivery using the APEX Analyze-Plan-Execute-Examine methodology.
SunS1eep1ngFit regression, GLM, and time series models in Python.
