kimun 0.13.3

Code metrics tool — health score, complexity, duplication, hotspots, ownership
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145

/// AI provider skill installer.
///
/// Installs a Claude Code skill file (`SKILL.md`) that teaches the AI
/// provider how to invoke the `km` CLI for code analysis. The skill
/// can be installed at project level (`.claude/skills/`) or user level
/// (`~/.claude/skills/`).
use std::fs;
use std::io::{self, Write};
use std::path::PathBuf;

/// Skill markdown content embedded at compile time.
/// Describes all available `km` subcommands and their JSON schemas.
const SKILL_CONTENT: &str = r#"---
name: km-analyze
description: Analyze a code repository using km (code metrics) tool
---

# km — Code Metrics Analysis Skill

You have access to the `km` CLI tool for comprehensive code analysis. Use it to analyze repositories across multiple dimensions.

## Available Commands

Run these via the Bash tool. Always use `--json` for machine-readable output.

### Lines of Code
```bash
km loc [PATH] --json
```
Language breakdown: files, blank lines, comment lines, code lines.

### Code Health Score
```bash
km score [PATH] --json
```
Overall grade (A++ to F--) across 6 dimensions: maintainability, complexity, duplication, indentation, Halstead effort, file size.

### Cyclomatic Complexity
```bash
km cycom [PATH] --json --top 20
```
Per-file and per-function complexity. High values indicate hard-to-test code.

### Maintainability Index
```bash
km miv [PATH] --json --top 20
```
Verifysoft variant (with comment weight). Values below 65 are hard to maintain.

### Halstead Complexity
```bash
km hal [PATH] --json --top 20 --sort-by effort
```
Effort, volume, and estimated bugs per file.

### Indentation Complexity
```bash
km indent [PATH] --json
```
Indentation depth stddev — high values suggest deeply nested code.

### Duplicate Code
```bash
km dups [PATH] --json --report
```
Duplicate blocks across the project.

### Hotspots (requires git)
```bash
km hotspots [PATH] --json --top 20
```
Files that change frequently AND have high complexity — top refactoring targets.

### Code Ownership (requires git)
```bash
km knowledge [PATH] --json --top 20
```
Bus factor risk per file via git blame analysis.

### Temporal Coupling (requires git)
```bash
km tc [PATH] --json --top 20
```
Files that change together in commits — hidden dependencies.

## Analysis Workflow

1. Start with `km score` for the overall health grade
2. Run `km loc` for project size and language breakdown
3. Use `km cycom` and `km miv` to find the most complex/unmaintainable files
4. Run `km hotspots` to find high-risk change-prone files
5. Check `km dups` for code duplication opportunities
6. Optionally run `km knowledge` and `km tc` for team/architecture insights

## Output Format

Produce a structured report with:
- **Overview**: Project size, languages, overall grade
- **Code Health**: Score breakdown by dimension
- **Complexity Hotspots**: Worst files by complexity
- **Maintainability**: Files hardest to maintain
- **Key Findings**: Notable patterns and risks
- **Recommendations**: Prioritized, actionable suggestions

Reference specific file names and metric values. Be concise but thorough.
"#;

/// Install the `km-analyze` skill for the given AI provider.
///
/// Prompts the user to choose between project-level and user-level
/// installation, creates the directory structure, and writes the skill file.
/// Currently only the `claude` provider is supported.
pub fn install(provider: &str) -> Result<(), Box<dyn std::error::Error>> {
    if provider != "claude" {
        return Err(format!("Unsupported provider: {provider}. Supported: claude").into());
    }

    println!("Where do you want to install the km skill?");
    println!("  1) Project-level (.claude/skills/km-analyze/)");
    println!("  2) User-level (~/.claude/skills/km-analyze/)");
    print!("Choose [1/2]: ");
    io::stdout().flush()?;

    let mut choice = String::new();
    io::stdin().read_line(&mut choice)?;
    let choice = choice.trim();

    let skill_dir = match choice {
        "1" => PathBuf::from(".claude/skills/km-analyze"),
        "2" => {
            let home = std::env::var("HOME").map_err(|_| "Could not determine home directory")?;
            PathBuf::from(home).join(".claude/skills/km-analyze")
        }
        _ => return Err("Invalid choice. Please enter 1 or 2.".into()),
    };

    fs::create_dir_all(&skill_dir)?;
    let skill_path = skill_dir.join("SKILL.md");
    fs::write(&skill_path, SKILL_CONTENT)?;

    println!("Skill installed at: {}", skill_path.display());
    println!("Claude Code will now be able to use km for code analysis.");

    Ok(())
}