Skip to content
Back to skills

Deps

ASecurity

Analyze .asmdef dependency relationships — Mermaid graph + dependency matrix + issue detection.

  • 11 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added May 26, 2026
development

Security analysis

A100/100

Scanned October 6, 2026

npx -y skills add tykisgod/quick-question --skill deps --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Deps?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Deps
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tykisgod-deps/badge)](https://www.skillsdirectory.com/skills/tykisgod-deps)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
description: "Analyze .asmdef dependency relationships — Mermaid graph + dependency matrix + issue detection."
---

Respond in the user's preferred language (detect from their recent messages, or fall back to the language setting in CLAUDE.md).

Arguments: $ARGUMENTS
- No arguments: analyze all .asmdef files in the project
- A module name (e.g. "Player"): analyze only that module's dependency chain, upstream and downstream; the Mermaid diagram shows only related modules

## Execution Steps

### 1. Collect All .asmdef Files

Use Glob to find `Assets/**/*.asmdef`, excluding Tests and Editor asmdefs (filenames containing `.Tests.`, `.Editor.`, or `.PlayModeTests.`).

If the project has no .asmdef files (pure directory-based project), scan `using` statements in .cs files instead, aggregate modules by namespace, and build the dependency graph from the using references.

### 2. Build GUID → Name Mapping

For each .asmdef, read its `name` field and the `guid` field of its `.asmdef.meta`.

### 3. Parse Dependency Relationships

For each .asmdef, map the GUIDs in its `references` array to names through the lookup table and record `ModuleA → [ModuleB, ModuleC]`.

### 4. Output Mermaid Dependency Graph

Generate a Mermaid flowchart, arranged by layer from top to bottom. Layers are inferred from the graph: modules with no dependencies are Layer 0, modules depending only on Layer 0 are Layer 1, etc.

````markdown
```mermaid
graph TD
    subgraph "Layer 0"
        Core
    end
    subgraph "Layer 1"
        Player
        AI
    end
    subgraph "Layer 2"
        UI
        Networking
    end

    Player --> Core
    AI --> Core
    UI --> Player
    Networking --> Core
```
````

- Normal dependencies: solid arrow `-->`
- Circular dependencies: red dashed `-.->|cycle|`
- Layer violations: orange bold `==>|violation|`

### 5. Output Dependency Matrix

```markdown
| Module ↓ Depends on → | Core | Player | AI | UI | ... |
|------------------------|:----:|:------:|:--:|:--:|:---:|
| Core                   |  -   |        |    |    |     |
| Player                 |  ✓   |   -    |    |    |     |
| AI                     |  ✓   |        | -  |    |     |
| UI                     |      |   ✓    |    | -  |     |
```

### 6. Detect Issues

**Circular dependencies:** run DFS on the dependency graph; list every full circular path.

**Layer violations** need declared layers: read them from `AGENTS.md` or the project's architecture docs and report every edge from a lower declared layer to a higher one. With no declared layers, show the inferred layering and cycles only, and say that no layer rule was checked.

### 7. Output Health Summary

```
### Dependency Health
- Total modules: N
- Average dependencies: X
- Most dependencies: ModuleName (Y dependencies)
- Circular dependencies: None / Found (list them)
- Layer violations: None / Found (list them) / Not checked (no declared layers)
```

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.