Evaluate existing code using the ganvil evaluation framework. Automatically determines if the project is frontend, backend, or fullstack and applies the appropriate evaluator with four-dimension scoring. Use to QA-check any existing project or codebase.
Installs into .claude/skills of the current project.
Are you the author of Evaluate?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/superduke-evaluate)
---
name: evaluate
description: >
Evaluate existing code using the ganvil evaluation framework. Automatically
determines if the project is frontend, backend, or fullstack and applies the
appropriate evaluator with four-dimension scoring. Use to QA-check any existing
project or codebase.
argument-hint: "[frontend|backend|auto] optional description of what to focus on"
---
# Ganvil Evaluate — Standalone QA Evaluation
Evaluate an existing project or codebase using the ganvil evaluation framework, without requiring a prior plan or build.
## What to Do
1. Determine the project type:
- If `$ARGUMENTS` starts with "frontend" or "backend", use that classification
- Otherwise, examine the project structure to auto-detect:
- Has `package.json` with frontend frameworks (React, Vue, Svelte, etc.) → FRONTEND
- Has backend files (Python, Go, Rust, Java server code, `requirements.txt`, `Cargo.toml`, etc.) → BACKEND
- Has both → evaluate as FULLSTACK (backend first, then frontend)
2. Create `ganvil-artifacts/` directory if it doesn't exist.
3. **For frontend evaluation**: Delegate to the **frontend-evaluator** agent.
- The evaluator will start the application, interact via browser, and score on:
- Design Quality (≥7 to pass)
- Originality (≥7 to pass)
- Craft (≥6 to pass)
- Functionality (≥7 to pass)
- Output: `ganvil-artifacts/frontend-evaluation.md`
4. **For backend evaluation**: Delegate to the **backend-evaluator** agent.
- The evaluator will start the server, run tests, hit API endpoints, and score on:
- API Correctness (≥7 to pass)
- Data Integrity (≥7 to pass)
- Code Quality (≥6 to pass)
- Functional Completeness (≥7 to pass)
- Output: `ganvil-artifacts/backend-evaluation.md`
5. **For fullstack evaluation**: Run both evaluators, backend first.
6. Summarize the evaluation results, highlighting:
- Overall pass/fail per dimension
- Critical bugs found
- Top recommendations for improvement
## Usage
```
# Auto-detect project type and evaluate
/ganvil:evaluate
# Evaluate only the frontend
/ganvil:evaluate frontend
# Evaluate with focus
/ganvil:evaluate frontend focus on mobile responsiveness and accessibility
# Evaluate backend
/ganvil:evaluate backend focus on API error handling
```
## Note
This skill evaluates code as-is. It does not require a `ganvil-artifacts/spec.md` or sprint contracts — the evaluator will assess the project against general quality standards. However, if a spec exists from a prior `/ganvil:plan`, the evaluator will use it for more targeted evaluation.