168 tools, each solving a specific problem.
That sounds powerful. In practice, your brain can’t hold all of them. You don’t remember what each one is called, which context it applies to, or when to reach for it. The result: you end up using the 5 you actually remember.
This is the paradox of abundance: more resources makes decisions harder.
The Problem Flipped
Claude Code Skills were designed to package best practices into reusable instruction sets — a good idea. But once the skill library grows past a certain size, the question stops being “is the right tool available?” and becomes “did I remember to use it?”
gstack-industrial flips the model: instead of you finding the right tool, the right tool finds you.
Three-Layer Routing
The core of gstack-industrial is a smart router that analyzes your message and current project state to proactively suggest the most relevant skill.
The analysis works across three layers:
- What you said: Keyword and semantic analysis. “Review” → suggest code-review. “Brainstorm” → suggest brainstorming.
- What stage you’re in: Git status and recently modified files infer what you’re working on.
- Contextual signals: Uncommitted changes, failing tests, open PRs.
Three signals combined produce more accurate suggestions than keyword matching alone.
Anti-Annoyance Is Not Optional
Suggestion systems have a failure mode: too many suggestions → users start ignoring them → the system becomes meaningless.
gstack-industrial addresses this directly: cooldown periods so the same suggestion doesn’t appear consecutively, per-session caps on total suggestions, and a “stop suggesting” command users can invoke at any time.
Suggestion quality and suggestion frequency both have to be right, independently.
The Template System
Skill libraries have a second problem beyond discoverability: update cost.
When you want to add a standard step across all your skills, you have to edit each one manually. Fifty skills means fifty edits.
gstack-industrial’s template system lets you define shared standards in one place. All skills inherit from the template. One update propagates everywhere.
From Toolset to Infrastructure
gstack-industrial is building something broader: transforming a skill library from a folder of documents into an intelligent infrastructure layer.
A developer’s personal skill library is a personal asset. An organization’s skill library is institutional knowledge. When that knowledge can’t be reliably surfaced at the moment it’s needed, it stays as inert files rather than active productivity.
The router solves the last mile: from “exists” to “gets used.”