skill-hub

تایید شده

OpenClaw skill discovery, security vetting & install. Searches 3000+ curated skills from ClawHub registry and awesome-openclaw-skills catalog. Scores credibility, detects prompt injection & malicious patterns, manages installations. Quick-checks GitHub for new skills.

@openclaw
v1.0.0MIT۱۴۰۴/۱۲/۳
62از ۱۰۰
(0)
۱.۰k
۲۰۹
۲۶۴

نصب مهارت

مهارت‌ها کدهای شخص ثالث از مخازن عمومی GitHub هستند. SkillHub الگوهای مخرب شناخته‌شده را اسکن می‌کند اما نمی‌تواند امنیت را تضمین کند. قبل از نصب، کد منبع را بررسی کنید.

نصب سراسری (سطح کاربر):

npx skillhub install openclaw/skills/skill-hub

نصب در پروژه فعلی:

npx skillhub install openclaw/skills/skill-hub --project

مسیر پیشنهادی: ~/.claude/skills/skill-hub/

بررسی هوش مصنوعی

کیفیت دستورالعمل65
دقت توضیحات60
کاربردی بودن57
صحت فنی72

Scored 62 — impressive automation with 7 real Python scripts including security vetting (NLP-level prompt injection detection). Strong technical execution. Limited by OpenClaw ecosystem lock-in — scripts reference ClawHub registry and specific catalog formats.

محتوای SKILL.md

---
name: skill-hub
description: "OpenClaw skill discovery, security vetting & install. Searches 3000+ curated skills from ClawHub registry and awesome-openclaw-skills catalog. Scores credibility, detects prompt injection & malicious patterns, manages installations. Quick-checks GitHub for new skills."
license: MIT
version: 1.0.0
homepage: https://github.com/PhenixStar/openclaw-skills-collection
user-invocable: true
disable-model-invocation: false
auto_activate:
  - "find skill"
  - "install skill"
  - "search skills"
  - "what skills exist"
  - "skill for"
  - "discover skill"
  - "vet skill"
  - "scan skill"
  - "skill security"
  - "new skills"
  - "skill updates"
  - "browse skills"
allowed-tools:
  - Bash
  - Read
  - Write
---

# Skill Hub

Unified skill discovery, security vetting, and installation for OpenClaw.

## Commands

### Search Skills
Find skills by keyword, category, or credibility score.

```bash
python3 scripts/skill-hub-search.py --query "spreadsheet"
python3 scripts/skill-hub-search.py --category "DevOps" --min-score 60
python3 scripts/skill-hub-search.py --query "auth" --live        # include live ClawHub results
python3 scripts/skill-hub-search.py --installed                  # show only installed
python3 scripts/skill-hub-search.py --not-installed --limit 20   # discovery mode
```

### Install Skills
After finding a skill, install via ClawHub:

```bash
npx clawhub@latest install <skill-slug>
```

### Vet Skills (Security Scan)
Scan a skill for malicious patterns, prompt injection, and logic weaknesses.

```bash
python3 scripts/skill-hub-vet.py --slug google-sheets     # vet single skill
python3 scripts/skill-hub-vet.py --all-installed           # vet all installed
python3 scripts/skill-hub-vet.py --category "DevOps"       # vet category
python3 scripts/skill-hub-vet.py --top 10                  # vet top N unvetted
```

### Status Dashboard
See installed vs catalog coverage, unvetted warnings, recommendations.

```bash
python3 scripts/skill-hub-status.py
```

### Quick Check (GitHub API)
Fast check if new skills were added since last sync. Uses `gh` CLI — no full download needed.

```bash
python3 scripts/skill-hub-quick-check.py              # check for updates
python3 scripts/skill-hub-quick-check.py --sync        # auto-sync if updates found
python3 scripts/skill-hub-quick-check.py --query "ai"  # check + search new skills
```

### Browse Full Catalog
Export catalog as formatted table (terminal or markdown), grouped by category.

```bash
python3 scripts/skill-hub-table-export.py                          # terminal table
python3 scripts/skill-hub-table-export.py --format markdown        # markdown table
python3 scripts/skill-hub-table-export.py --category "AI"          # filter category
```

### Sync Catalog
Full re-fetch from GitHub awesome-list. Computes credibility, preserves vet results, shows diff.

```bash
python3 scripts/skill-hub-sync.py
```

## Credibility Scores (0-100)

| Tier | Score | Meaning |
|------|-------|---------|
| Trusted | 85-100 | Curated + vetted + mature |
| Good | 60-84 | Curated or vetted, some signals |
| Unvetted | 30-59 | Exists in registry, not scanned |
| Caution | 0-29 | Missing signals or security warnings |

## Security Checks

Code-level: eval/exec, shell injection, obfuscation, network access, env harvesting, destructive ops.

NLP/Prompt-level: hidden instructions, role hijacking, invisible unicode, exfiltration prompts, authority escalation, social engineering.

## When to Use

- User asks "find a skill for X" or "is there a skill that can..."
- User wants to extend capabilities with new tools
- User wants to check if installed skills are safe
- Before installing unknown skills from registry