v0.1.1

alon-fact-check

SuspiciousClawScan verdict for this skill. Analyzed Apr 30, 2026, 3:50 PM.

Analysis

The fact-checking instructions are mostly appropriate, but the skill advertises crypto and purchase-related capabilities that are not needed for fact-checking.

GuidanceInstall only if you need web-based fact-checking, avoid submitting confidential material, verify the install source if using the README command, and do not grant crypto or purchase permissions for this skill unless the publisher clearly removes or justifies those capabilities.

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

SKILL.md

For each claim, perform searches ... Use the host's available search and browsing tools. Open and inspect sources before citing them

The skill relies on browsing and search tools, which is appropriate for fact-checking, but users should expect external web lookups.

User impactThe agent may browse URLs and search the web while checking claims.

RecommendationUse this with information you are comfortable sending through the host's browsing or search tools.

Agentic Supply Chain Vulnerabilities

SeverityLowConfidenceMediumStatusNote

README.md

npx skills add alondotsh/alon-skills --skill alon-fact-check

The README install guidance uses npx and an external repository reference without a pinned version; the submitted artifact itself has no install spec or code files.

User impactUsers following the README may depend on whatever the external repository and package tool provide at install time.

RecommendationVerify the repository and prefer pinned versions or registry-provided installation paths.

Permission boundary

Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.

Identity and Privilege Abuse

SeverityHighConfidenceHighStatusConcern

metadata

Description: USE WHEN user wants to verify factual claims ... Capability signals: crypto; can-make-purchases

Crypto and purchase authority are not needed for the stated fact-checking purpose, and the artifacts do not bound approval, scope, or reversibility for those high-impact capabilities.

User impactIf a platform grants these capabilities, the skill may receive more authority than a fact-checker needs.

RecommendationDo not grant crypto or purchase-related permissions for this skill unless the publisher removes, justifies, and tightly scopes them.

Sensitive data protection

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

Insecure Inter-Agent Communication

SeverityLowConfidenceMediumStatusNote

SKILL.md

use the host's available web reader, browser, or search tools to fetch readable page content

The skill may send user-provided URLs, pasted claims, or derived queries through host-provided web/search tools; the specific provider boundaries are not described.

User impactPrivate or sensitive claims could be exposed to external browsing or search providers during verification.

RecommendationAvoid submitting confidential text or private URLs unless you trust the host's browsing and search data handling.