url-eval
Evaluates candidate domain names/URLs for a website. Takes a site description and a list of candidate URLs, scores each across 8 research-ba
Open source Open in the app JSON README (API)
About
Evaluates candidate domain names/URLs for a website. Takes a site description and a list of candidate URLs, scores each across 8 research-backed dimensions, checks availability via web search, suggests alternatives, and produces a structured Markdown report. Invoke as /url-eval.
Details
- Kind
- Plugins
- Topic
- Files & documents
- Publisher
- keithmackay
- Origin
- gemini
- Category
- ferramentas
- Version
- 1.0.0
- Stars
- 1
- Last push
- 2026-08-22T19:21:28Z
- Repository state
- ativo
- Language
- Shell
- License
- MIT
- Added
- 2026-08-30 14:13:39
- Updated
- 2026-08-30 14:13:39
- Origin id
keithmackay/url-eval
README
# url-eval
A Claude Code skill that evaluates candidate domain names for a website — scoring them across 8 research-backed dimensions, checking availability, and suggesting alternatives you may not have considered.
## What it does
Upload a site description and a list of candidate URLs, and `/url-eval` produces a structured Markdown report with:
- **Top 3 recommendations** — score, narrative, availability status
- **Candidate narratives** — plain-English analysis of each domain you submitted
- **Full score table** — all candidates + alternatives, sorted by overall score
Scoring criteria are derived from academic and industry research on domain name memorability, spelling reliability, brand fit, and more. The `--update` flag triggers a live web search to supplement the baked-in research with current findings before scoring.
## Installation
### From the mackayi marketplace (recommended)
```
/plugin marketplace add keithmackay/mackayi
/plugin install url-eval@mackayi
```
### Claude Code
```bash
git clone https://github.com/keithmackay/url-eval ~/.claude/plugins/url-eval
cd ~/.claude/plugins/url-eval
mkdir -p ~/.claude/skills/url-eval
ln -sf "$(pwd)/SKILL.md" ~/.claude/skills/url-eval/SKILL.md
```
Restart Claude Code, type `/`, and confirm `url-eval` appears in the autocomplete list.
### Codex
Add an entry to your marketplace JSON (`~/.agents/plugins/marketplace.json`, create if absent):
```json
{
"name": "personal",
"interface": { "displayName": "Personal Plugins" },
"plugins": [
{
"name": "url-eval",
"source": { "source": "local", "path": "/path/to/url-eval/" },
"policy": { "installation": "AVAILABLE", "authentication": "ON_INSTALL" },
"category": "Productivity"
}
]
}
```
Then invoke with `/url-eval`.
### Antigravity
The root `SKILL.md` is natively compatible — no conversion needed.
**Global install** (all workspaces):
```bash
cp -r /path/to/url-eval/ ~/.gemini/antigravity/skills/url-eval/
```
**Workspace install** (current project only):
```bash
cp -r /path/to/url-eval/ .agents/skills/url-eval/
```
Skills are auto-discovered. You can also invoke by name: `/url-eval`.
### Gemini CLI
Gemini CLI installs extensions directly from GitHub:
```bash
gemini extensions install https://github.com/keithmackay/url-eval
```
To update:
```bash
gemini extensions update url-eval
```
The skill is auto-discovered from `GEMINI.md` after installation.
## Usage
### Interactive mode (recommended)
```
/url-eval
```
Claude will ask for your site description, then your candidate URLs — one question at a time.
### Inline mode
```
/url-eval --site "description of your site" candidate1.com candidate2.io
```
### With research refresh
```
/url-eval --update --site "description" candidate1.com
```
Runs live web searches for recent domain naming research before scoring. Changes apply to the current session only; see [Updating the research baseline](#updating-the-research-baseline) to make them permanent.
## Scoring dimensions
| Dimension | Weight | What it measures |
|------------------------|--------|-----------------|
| Memorability | 20% | Length, syllables, distinctiveness |
| Spelling Reliability | 15% | Typo risk, homophones, hyphens |
| Pronunciation Clarity | 10% | Ambiguity, silent letters |
| Associations | 15% | Positive/negative connotations, cultural fit |
| Cleverness / Brand Fit | 10% | Uniqueness, trademark potential |
| Relevance to Purpose | 15% | How well the name signals site content |
| Competitor Overlap | 10% | Similarity to existing brands |
| TLD Appropriateness | 5% | TLD trust and context fit |
**Overall score: 20–100.** Scores above 70 are strong candidates.
## Research foundation
Scoring criteria are drawn from a curated research summary in [`docs/research/DOMAIN_NAME_RESEARCH.md`](docs/research/DOMAIN_NAME_RESEARCH.md), which covers 30+ academic and industry sources on memorability, spelling reliability, TLD trust, brand distinctiveness, and more.
Run `/url-eval --update` to supplement the baked-in summary with live web searches for current findings. The skill will produce an Update Report at the top of the response describing what, if anything, changed.
## Updating the research baseline
To make research updates permanent:
1. Run `/url-eval --update` and note the findings
2. Edit [`docs/research/DOMAIN_NAME_RESEARCH.md`](docs/research/DOMAIN_NAME_RESEARCH.md) with new findings
3. Update the `<!-- RESEARCH_SUMMARY_START -->` section in `SKILL.md` to match
4. If installed via symlink (recommended): no action needed — the symlink always points to the latest file. If installed via copy: `cp SKILL.md ~/.claude/skills/url-eval/SKILL.md`
5. Commit and push
## Development
Manual test cases covering all 8 phases of the skill are in [`docs/testing/TEST_CASES.md`](docs/testing/TEST_CASES.md). There is no automated test runner — tests are invoked manually inside Claude Code.
To run a test:
```
/url-eval --site "recipe sharing platform for home cooks" recipebox.com homechef.io cookwith.me
```
Compare the output against the expected behavior documented in `TEST_CASES.md`. Record actual outputs in [`docs/testing/SAMPLE_RUNS.md`](docs/testing/SAMPLE_RUNS.md) for regression reference.
To contribute a change, see [CONTRIBUTING.md](CONTRIBUTING.md).
## Compatibility
| Feature | Claude Code | Codex | Antigravity | Gemini CLI |
|---------|:-----------:|:-----:|:-----------:|:----------:|
| Core skill | ✅ | ✅ | ✅ | ✅ |
| Interactive input prompting | ✅ | ✅ | ✅ | ✅ |
| Inline mode (`--site`, URL args) | ✅ | ✅ | ✅ | ✅ |
| `--update` flag (web search) | ✅ | ✅ | ✅ | ✅ |
| 8-dimension scoring | ✅ | ✅ | ✅ | ✅ |
| Availability checking (web search) | ✅ | ✅ | ✅ | ✅ |
| Alternative suggestions | ✅ | ✅ | ✅ | ✅ |
Legend: ✅ Supported
All features are fully portable — the skill uses no platform-specific metadata, triggers, or tool APIs beyond standard web search.
## References
- **Claude Code Skills:** https://code.claude.com/docs/en/skills
- **Codex Plugins:** https://developers.openai.com/codex/plugins/build
- **Antigravity Skills:** https://antigravity.google/docs/skills
- **Gemini CLI Extensions:** https://github.com/google-gemini/gemini-cli/blob/main/docs/extension.md
- **Agent Skills open standard:** https://agentskills.io/home
## Changelog
See [CHANGELOG.md](CHANGELOG.md) for release history.
## License
[MIT](LICENSE)