# cyberlens

Adopt this skill: https://proskills.md/skills/other/shadoprizm-cyberlens-skill

Page: https://proskills.md/skills/other/shadoprizm-cyberlens-skill
This document is the public skill bundle for that page. It contains SKILL.md and every helper file the catalog carries. No login is required.

## File: SKILL.md
size: 7994
upstream: https://github.com/shadoprizm/cyberlens-skill/blob/HEAD/SKILL.md

````text
---
name: cyberlens
description: Scan websites, GitHub repositories, and Claw Hub skills for practical security issues using a local quick website scan and CyberLens cloud analysis when connected.
metadata: {"openclaw": {"requires": {"bins": ["python3"]}, "primaryEnv": "CYBERLENS_API_KEY", "emoji": "🔒", "homepage": "https://cyberlensai.com"}}
---

# CyberLens Security Scanner

Scan websites, GitHub repositories, and Claw Hub skills for practical security issues before you ship, install, or trust them. CyberLens can audit live web targets for missing headers, HTTPS weaknesses, exposed technologies, and insecure forms. It can scan GitHub repositories for dependency vulnerabilities, secret leaks, and malicious code. And it can scan Claw Hub skills directly from their Claw Hub URL — just paste a link like `https://clawhub.ai/author/skill-name` and CyberLens will analyse the skill package before you install it. Results include a 0-100 security score, letter grade (A-F), AI-powered analysis, and plain-English remediation advice. Reports can be shared as formatted markdown in chat or exported as a PDF.

## Prerequisites

- **Python 3.9+**: Required for script execution
- **Python Packages**:
  ```bash
  pip install -r requirements.txt
  ```

## First-Time Setup

Website quick scans and Claw Hub skill package scans work without an account. Connect a CyberLens account when the user wants the full cloud website scan or any repository scan. Run the `connect_account` tool. This opens a browser to cyberlensai.com where they sign in or create an account. A short-lived connect code is delivered through the callback, the skill exchanges that code for the real account key over HTTPS on official CyberLens hosts, and the key is stored at `~/.openclaw/skills/cyberlens/config.yaml`.

Browser authentication uses `https://cyberlensai.com/connect`. The hosted scan API runs at `https://api.cyberlensai.com/functions/v1/public-api-scan`. If the user needs to override the scan API endpoint explicitly, set `CYBERLENS_API_BASE_URL`.

If the user doesn't have a CyberLens account, direct them to https://cyberlensai.com to sign up. Free tier includes 5 scans/month (3 website + 2 repository). Repository scanning requires a connected account. If the user exhausts quota, the skill opens the pricing page and returns an upgrade URL in the tool result.

## Tools

### connect_account

Connect or reconnect a CyberLens account. Opens the browser to cyberlensai.com/connect, waits for authentication, validates the returned exchange host against official CyberLens infrastructure, and stores the API key locally.

```bash
python3 -c "
import asyncio
from cyberlens_skill.src.tools import connect_account
result = asyncio.run(connect_account())
print(result)
"
```

Use this when:
- The user wants to connect their CyberLens account
- The user gets an authentication error during scanning
- The user wants to switch to a different account

If OpenClaw is running on another machine, set `CYBERLENS_CONNECT_CALLBACK_URL` to a browser-reachable callback URL before running `connect_account`. Use `CYBERLENS_CONNECT_BIND_HOST` and `CYBERLENS_CONNECT_BIND_PORT` when a reverse proxy or different local bind address is involved. If no callback path is available, the user can still set `CYBERLENS_API_KEY` manually. If the user prefers not to persist the key on disk, they can keep `CYBERLENS_API_KEY` only in the process environment.

### scan_target

Scan a live website, GitHub repository, or Claw Hub skill URL. CyberLens auto-detects the target type. Website scans use the local quick engine without an account and the full cloud path when connected. Repository scans require the CyberLens cloud API.

Parameters:
- `target` (required): Website URL, GitHub repository URL, or Claw Hub skill URL
- `scan_depth`: requested depth: "quick", "standard" (default), or "deep". Local website mode always uses the quick scan and warns when a deeper mode was requested.
- `timeout`: Request timeout in seconds (default: 30)
- `use_cloud`: Force cloud (true) or local (false) scanning. Repository and skill scans require cloud.

```bash
python3 -c "
import asyncio
from cyberlens_skill.src.tools import scan_target
result = asyncio.run(scan_target('https://clawhub.ai/anthropic/tavily-search'))
print(result)
"
```

### scan_skill

Scan a Claw Hub skill before installing it. Accepts a Claw Hub URL (e.g. `https://clawhub.ai/author/skill-name`) or a direct skill download URL. The skill package is downloaded and analysed locally for security vulnerabilities, malicious code, dependency issues, secret leaks, and trust posture problems.

Use this when:
- A user wants to check a Claw Hub skill before running `clawhub install`
- A user pastes a Claw Hub link and asks whether the skill is safe
- You want to verify a skill's security posture before recommending installation

Parameters:
- `skill_url` (required): Claw Hub skill URL or direct skill download URL
- `timeout`: Request timeout in seconds (default: 60)

```bash
python3 -c "
import asyncio
from cyberlens_skill.src.tools import scan_skill
result = asyncio.run(scan_skill('https://clawhub.ai/anthropic/tavily-search'))
print(result)
"
```

Returns: security_score, grade (A-F), local assessment, and detailed findings with file paths and recommendations.

### scan_website

Scan a website URL for security vulnerabilities. Uses the local quick scan without an account and the CyberLens cloud API when connected (70+ checks). For repository or Claw Hub URLs, use `scan_target`, `scan_skill`, or `scan_repository`.

Parameters:
- `url` (required): The website URL to scan (must include https:// or http://)
- `scan_depth`: requested depth: "quick", "standard" (default), or "deep". Local website mode always uses the quick scan and warns when a deeper mode was requested.
- `timeout`: Request timeout in seconds (default: 30)
- `use_cloud`: Force cloud (true) or local (false) scanning. Auto-detects by default.

```bash
python3 -c "
import asyncio
from cyberlens_skill.src.tools import scan_website
result = asyncio.run(scan_website('https://example.com'))
print(result)
"
```

Returns: score (0-100), grade (A-F), findings with severity/description/remediation, and scan source (cloud or local).

### scan_repository

Scan a GitHub repository URL, including OpenClaw skills before installation. Repository scans use the CyberLens cloud API and return repository findings, dependency alerts, trust posture findings, and security/trust scores.

Parameters:
- `repository_url` (required): GitHub repository URL such as `https://github.com/owner/repo`
- `timeout`: Request timeout in seconds (default: 60)
- `use_cloud`: Force cloud behaviour. Repository scans require cloud access.

```bash
python3 -c "
import asyncio
from cyberlens_skill.src.tools import scan_repository
result = asyncio.run(scan_repository('https://github.com/shadoprizm/cyberlens-skill'))
print(result)
"
```

### get_security_score

Quick security score check. Faster than a full scan when you only need the grade. Works with website URLs, GitHub repositories, and Claw Hub skill URLs.

Parameters:
- `url` (required): The URL to check (website, GitHub repo, or Claw Hub skill)
- `timeout`: Request timeout in seconds (default: 30)

```bash
python3 -c "
import asyncio
from cyberlens_skill.src.tools import get_security_score
result = asyncio.run(get_security_score('https://clawhub.ai/anthropic/tavily-search'))
print(result)
"
```

### generate_report

Generate a formatted markdown report from any scan result. Takes the output of `scan_target`, `scan_skill`, `scan_website`, or `scan_repository` and produces a clean markdown document suitable for sharing via messaging (Telegram, Discord, Signal), the web UI, or any channel that renders markdown.

Use this when:
- The user asks to see a report or summary of the scan
- The user wants to share results in a chat or messaging platform
- You want to present findings in a readable format

Parameters:
- `scan_result
````

## File: README.md
upstream: https://github.com/shadoprizm/cyberlens-skill/blob/HEAD/README.md

````text
---
name: cyberlens-security-scanner
description: Scan websites, GitHub repositories, and Claw Hub skills for security vulnerabilities before you ship, install, or trust them. Supports live web targets (headers, HTTPS, exposed tech, insecure forms), GitHub repos (dependencies, secrets, malicious code), and Claw Hub skill packages (trust posture, behavioural analysis). Returns a 0-100 security score, letter grade (A-F), AI-powered analysis, and plain-English remediation. Export results as formatted markdown or PDF. Use when the user wants to audit a URL, check a skill before installing, scan a repo for vulnerabilities, or generate a security report.
---

# CyberLens Security Scanner

[![Open CLAW](https://img.shields.io/badge/Open%20CLAW-Native%20Skill-blue)](https://openclaw.ai)
[![Claw Hub](https://img.shields.io/badge/Claw%20Hub-Available-brightgreen)](https://clawhub.ai)
[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-green.svg)](LICENSE)
[![Python](https://img.shields.io/badge/Python-3.9%2B-3776AB)](https://www.python.org/downloads/)
[![CyberLens](https://img.shields.io/badge/CyberLens-Powered-orange)](https://cyberlensai.com)

Stop guessing whether that website, repo, or skill is safe. CyberLens gives you a security score, a letter grade, and clear remediation steps in seconds.

## What You Get

**Website scanning** has two honest modes: a **local quick scan** without an account and a **full cloud scan** when connected. The local quick scan covers ~15 core checks; the cloud scan runs 70+ checks with richer analysis.

**Repository scanning** analyses GitHub repos for dependency vulnerabilities, leaked secrets, malicious code patterns, and trust posture issues.

**Claw Hub skill scanning** downloads the skill package, extracts it, and analyses the actual source code for dangerous patterns, hardcoded secrets, permission issues, and dependency risks. Paste a Claw Hub URL or a direct download link and CyberLens handles the rest. No account required for skill scanning.

Every scan returns a **0-100 security score**, an **A-F letter grade**, and **plain-English remediation advice**. Results can be viewed as formatted markdown in chat or exported as a professional PDF.

## Quick Start

```
clawhub install cyberlens-security-scanner
```

### 1. Scan something

Just tell your agent what to scan:

- "Check this skill before I install it: https://clawhub.ai/author/skill-name"
- "Scan https://example.com for security issues"
- "Audit https://github.com/owner/repo for vulnerabilities"
- "Give me a quick security score for https://example.com"

CyberLens auto-detects the target type and selects the right scan method. Website quick scans and Claw Hub skill scans work immediately with no account. Repository scans require a connected account.

### 2. Connect your account (for cloud scans)

Without an account, website scans use the local quick scanner and repository scans stay unavailable. Connect your account to upgrade website scans to the full cloud path and unlock repository scanning. The `connect_account` tool opens a browser, completes authentication, and stores your API key locally. Free accounts get 5 cloud scans/month (3 website + 2 repository). No credit card required. If you exhaust website cloud quota, the skill opens the CyberLens pricing page and falls back to the local quick scan automatically.

Don't have an account? Sign up at [cyberlensai.com](https://cyberlensai.com).

### 3. Get your report

After any scan, choose how you want results delivered:

- **In chat** via `generate_report` (formatted markdown)
- **As a PDF** via `export_report_pdf` (Security Assessment Dossier with an executive assessment, severity ledger, and complete technical evidence)

---

## Tools

### scan_skill

Scan a Claw Hub skill before installing it. This is the primary tool for vetting OpenClaw skills from the Claw Hub marketplace.

The skill package is downloaded, extracted, and analysed locally -- no CyberLens account required. Accepts Claw Hub URLs (the download link is resolved automatically) or direct download URLs.

**Parameters:**
- `skill_url` (required) -- Claw Hub skill URL (e.g. `https://clawhub.ai/skills/skill-name`) or direct download URL
- `timeout` -- Request timeout in seconds (default: 60)

**Example prompts:**
- "Scan https://clawhub.ai/skills/ontology before I install it"
- "Is this Claw Hub skill safe? https://clawhub.ai/author/skill-name"
- "Check this skill for malicious code before I install it"

**Returns:** Security score, grade (A-F), assessment, and detailed findings with file locations (dangerous code patterns, hardcoded secrets, manifest issues, dependency problems, external URLs).

### scan_target

Scan a live website, GitHub repository, or Claw Hub skill URL. CyberLens auto-detects the target type.

**Parameters:**
- `target` (required) -- Website URL, GitHub repository URL, or Claw Hub skill URL
- `scan_depth` -- Requested depth: `"quick"`, `"standard"` (default), or `"deep"`. Local website mode always uses the quick scan and warns when a deeper mode was requested.
- `timeout` -- Request timeout in seconds (default: 30)
- `use_cloud` -- Force cloud (`true`) or local (`false`). Repository scans require cloud.

**Example prompts:**
- "Scan https://clawhub.ai/author/skill-name for security issues"
- "Scan https://example.com for security issues"
- "Scan https://github.com/shadoprizm/cyberlens-skill before I install it"

### connect_account

Connect or reconnect your CyberLens account for cloud-powered scanning.

**Example prompt:** "Connect my CyberLens account"

### scan_website

Scan a website for security vulnerabilities. Uses the local quick engine without an account and the full cloud API when connected.

**Parameters:**
- `url` (required) -- The URL to scan (must include `https://` or `http://`)
- `scan_depth` -- Requested depth: `"quick"`, `"standard"` (default), or `"deep"`. Local website mode always uses the quick scan and warns when a deeper mode was requested.
- `timeout` -- Request timeout in seconds (default: 30)
- `use_cloud` -- Force cloud (`true`) or local (`false`). Auto-detects by default.

**Example prompts:**
- "Scan https://example.com for security issues"
- "Run the full cloud website scan for https://mysite.com"
- "Check if https://example.com is secure"

**Returns:** Score (0-100), grade (A-F), findings with severity/description/remediation, scan source (cloud or local), and explicit mode metadata showing whether this was a local quick scan or the full cloud scan.

### scan_repository

Scan a GitHub repository for security issues. Requires a connected CyberLens account.

**Parameters:**
- `repository_url` (required) -- GitHub repository URL (e.g. `https://github.com/owner/repo`)
- `timeout` -- Request timeout in seconds (default: 60)
- `use_cloud` -- Force cloud (`true`) or local (`false`). Repository scans require cloud access.

**Example prompts:**
- "Scan https://github.com/shadoprizm/cyberlens-skill for repo vulnerabilities"
- "Audit this OpenClaw skill before I install it"
- "Check my GitHub repo for security issues"

**Returns:** Repository security score, trust score, aggregated findings, and the underlying repository assessment sections from CyberLens cloud analysis.

### generate_report

Generate a formatted markdown report from any scan result. Suitable for sharing in Telegram, Discord, Signal, the web UI, or any channel that renders markdown.

**Parameters:**
- `scan_result` (required) -- The result returned by any CyberLens scan tool

**Example prompts:**
- "Show me the report in chat"
- "Give me a summary of the scan results"
- "Share the findings here"

**Returns:** A clean markdown report with score cards, AI analysis, summary, and severity-sorted findings.

### export_report_pdf

Export scan results as a professionally formatted PDF file with colour-coded severity indicators and full findings detail.

**Parameters:**
- `scan_result` (required) -- The result returned by any Cyb
````
