v1.0.0

Web Crawl

BenignClawScan verdict for this skill. Analyzed May 1, 2026, 7:42 AM.

Analysis

This appears to be a coherent web-crawling research helper, but it can run local Python and fetch arbitrary web pages, so users should review URLs and dependencies before use.

GuidanceThis skill is reasonable for web research. Before installing or invoking it, be comfortable with installing the Python dependencies, running the included local crawler code, and allowing the agent to fetch web pages for your research task. Avoid crawling sensitive internal pages unless intentional, and treat retrieved webpage text as evidence rather than instructions.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

Abnormal behavior control

Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.

Tool Misuse and Exploitation

SeverityLowConfidenceHighStatusNote

web_crawl.py

resp = requests.get(url, headers=self.headers, timeout=self.timeout, allow_redirects=True)

The crawler can request caller-supplied URLs and follow redirects. This is expected for a web crawler, but it means URL choice matters.

User impactIf the agent is given or discovers an unintended URL, it may fetch that destination from the user's environment.

RecommendationUse only intended HTTP/HTTPS URLs, avoid private/internal targets unless explicitly desired, and consider adding URL scheme/host validation for safer deployments.

Agentic Supply Chain Vulnerabilities

SeverityLowConfidenceHighStatusNote

README.md

pip3 install requests beautifulsoup4

The skill documents manual installation of unpinned Python dependencies, while the registry lists no install spec. The dependencies are expected for this crawler but still affect supply-chain review.

User impactInstalling the dependencies may pull the latest package versions from the package index rather than a pinned, reproducible set.

RecommendationInstall dependencies from trusted sources, pin versions where possible, and add an explicit install specification if this skill is distributed for repeatable installs.

Unexpected Code Execution

SeverityLowConfidenceHighStatusNote

EXAMPLES.md

exec:1 { "command": "cd ~/.openclaw/workspace-main/skills/web-crawl && python3 -c ... parallel_crawl(...)" }

The examples show using an exec command to run the local Python crawler. This is disclosed and central to the skill, but it is broader than a scoped tool call.

User impactThe skill may involve local Python execution, so command contents and pasted URL values should be reviewed.

RecommendationPrefer a scoped tool/function invocation when available; if using exec, review the command and ensure any inserted URLs or arguments are safely quoted.

Sensitive data protection

Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.

Memory and Context Poisoning

SeverityLowConfidenceHighStatusNote

EXAMPLES.md

Use the crawled content to:
- Extract key findings
- Compare sources
- Identify unique insights
- Cite sources

The skill places untrusted webpage content into the agent's analysis context. That is expected for research, but pages can contain instructions or misleading text.

User impactA malicious or unreliable page could try to influence the agent's reasoning or output if treated as instructions instead of source material.

RecommendationTreat crawled pages as untrusted evidence, keep source boundaries clear, and ignore any webpage text that attempts to direct the agent's behavior.