cc-reviewer 1.1.0 → 1.1.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +131 -0
- package/package.json +1 -1
package/README.md
ADDED
|
@@ -0,0 +1,131 @@
|
|
|
1
|
+
# CC Reviewer - AI Code Review for Claude Code
|
|
2
|
+
|
|
3
|
+
Get second-opinion feedback from OpenAI Codex and Google Gemini CLIs on Claude Code's work, then synthesize and incorporate.
|
|
4
|
+
|
|
5
|
+
## Quick Install
|
|
6
|
+
|
|
7
|
+
```bash
|
|
8
|
+
claude mcp add -s user cc-reviewer -- npx -y cc-reviewer
|
|
9
|
+
```
|
|
10
|
+
|
|
11
|
+
That's it! Restart Claude Code and the tools are available.
|
|
12
|
+
|
|
13
|
+
Verify with:
|
|
14
|
+
```bash
|
|
15
|
+
claude mcp list
|
|
16
|
+
# cc-reviewer: npx -y cc-reviewer - ✓ Connected
|
|
17
|
+
```
|
|
18
|
+
|
|
19
|
+
### Alternative: Manual Install
|
|
20
|
+
|
|
21
|
+
```bash
|
|
22
|
+
git clone https://github.com/SimonRen/cc-reviewer.git
|
|
23
|
+
cd cc-reviewer/mcp-server
|
|
24
|
+
npm install && npm run build
|
|
25
|
+
claude mcp add -s user cc-reviewer -- node /path/to/cc-reviewer/mcp-server/dist/index.js
|
|
26
|
+
```
|
|
27
|
+
|
|
28
|
+
## Prerequisites
|
|
29
|
+
|
|
30
|
+
Install at least one AI CLI:
|
|
31
|
+
|
|
32
|
+
```bash
|
|
33
|
+
# OpenAI Codex CLI
|
|
34
|
+
npm install -g @openai/codex-cli
|
|
35
|
+
codex login
|
|
36
|
+
|
|
37
|
+
# Google Gemini CLI
|
|
38
|
+
npm install -g @google/gemini-cli
|
|
39
|
+
gemini # follow auth prompts
|
|
40
|
+
```
|
|
41
|
+
|
|
42
|
+
## Slash Commands (Optional)
|
|
43
|
+
|
|
44
|
+
Copy the slash commands to your global commands folder:
|
|
45
|
+
|
|
46
|
+
```bash
|
|
47
|
+
# Clone repo (if not already)
|
|
48
|
+
git clone https://github.com/SimonRen/cc-reviewer.git
|
|
49
|
+
|
|
50
|
+
# Copy commands
|
|
51
|
+
cp cc-reviewer/commands/*.md ~/.claude/commands/
|
|
52
|
+
```
|
|
53
|
+
|
|
54
|
+
Then use:
|
|
55
|
+
|
|
56
|
+
```bash
|
|
57
|
+
/codex-review # Review with Codex
|
|
58
|
+
/codex-review security # Focus on security
|
|
59
|
+
|
|
60
|
+
/gemini-review # Review with Gemini
|
|
61
|
+
/gemini-review architecture # Focus on architecture
|
|
62
|
+
|
|
63
|
+
/multi-review # Both models in parallel
|
|
64
|
+
```
|
|
65
|
+
|
|
66
|
+
## How It Works
|
|
67
|
+
|
|
68
|
+
```
|
|
69
|
+
CC does work → User: /codex-review → External CLI reviews → CC synthesizes → Updated output
|
|
70
|
+
```
|
|
71
|
+
|
|
72
|
+
**Key Principles:**
|
|
73
|
+
- **CC is primary**: Claude Code does all the work; external models only review
|
|
74
|
+
- **Working directory strategy**: Pass `cwd` + small CC output; external CLIs read files directly
|
|
75
|
+
- **Synthesis, not passthrough**: CC always judges external feedback before incorporating
|
|
76
|
+
|
|
77
|
+
## Focus Areas
|
|
78
|
+
|
|
79
|
+
| Area | Description |
|
|
80
|
+
|------|-------------|
|
|
81
|
+
| `security` | Vulnerabilities, auth, input validation |
|
|
82
|
+
| `performance` | Speed, memory, efficiency |
|
|
83
|
+
| `architecture` | Design patterns, structure, coupling |
|
|
84
|
+
| `correctness` | Logic errors, edge cases, bugs |
|
|
85
|
+
| `maintainability` | Code clarity, documentation, complexity |
|
|
86
|
+
| `scalability` | Load handling, bottlenecks |
|
|
87
|
+
| `testing` | Test coverage, test quality |
|
|
88
|
+
| `documentation` | Comments, docs, API docs |
|
|
89
|
+
|
|
90
|
+
## MCP Tools
|
|
91
|
+
|
|
92
|
+
The plugin exposes four MCP tools:
|
|
93
|
+
|
|
94
|
+
| Tool | Description |
|
|
95
|
+
|------|-------------|
|
|
96
|
+
| `codex_feedback` | Get Codex review (correctness, edge cases, performance) |
|
|
97
|
+
| `gemini_feedback` | Get Gemini review (design patterns, scalability, tech debt) |
|
|
98
|
+
| `multi_feedback` | Parallel review from both models |
|
|
99
|
+
| `council_feedback` | Multi-model consensus with verification pipeline |
|
|
100
|
+
|
|
101
|
+
## Output Format
|
|
102
|
+
|
|
103
|
+
External CLIs return structured JSON feedback with:
|
|
104
|
+
- **Findings**: Issues with severity, confidence, location, and suggestions
|
|
105
|
+
- **Agreements**: Validations of CC's correct assessments
|
|
106
|
+
- **Disagreements**: Challenges to CC's claims with corrections
|
|
107
|
+
- **Alternatives**: Different approaches with tradeoffs
|
|
108
|
+
- **Risk Assessment**: Overall risk level with top concerns
|
|
109
|
+
|
|
110
|
+
## Development
|
|
111
|
+
|
|
112
|
+
```bash
|
|
113
|
+
cd mcp-server
|
|
114
|
+
npm install
|
|
115
|
+
npm run build # Build once
|
|
116
|
+
npm run dev # Watch mode
|
|
117
|
+
npm test # Run tests
|
|
118
|
+
npm run test:watch # Watch mode tests
|
|
119
|
+
npm start # Run server
|
|
120
|
+
```
|
|
121
|
+
|
|
122
|
+
## Publishing
|
|
123
|
+
|
|
124
|
+
Uses npm Trusted Publishing (OIDC, no tokens):
|
|
125
|
+
```bash
|
|
126
|
+
gh workflow run publish.yml -f version=patch # or minor/major
|
|
127
|
+
```
|
|
128
|
+
|
|
129
|
+
## License
|
|
130
|
+
|
|
131
|
+
MIT
|