T09 · Insecure Skill Coding Practices
- Location
SKILL.md:41- Finding
Unrestricted URL Fetching Enables Server-Side Request Forgery
- Content
View full analysis
- Remediation
View remediation
Security audit
Security checks for vulnerabilities and agentic risk
This blog-monitoring skill is coherent, but it fetches arbitrary user-provided URLs without network safety limits, so users should review it before installing.
Install only if you trust the environment and will restrict monitored URLs to public blog or RSS endpoints. Avoid localhost, private-network, cloud metadata, or other sensitive URLs, and prefer adding URL validation, request timeouts, response-size limits, exception handling, and pinned dependency versions before use.
SKILL.md:41Unrestricted URL Fetching Enables Server-Side Request Forgery
SKILL.md:62Unbounded Network Requests Can Cause Resource Exhaustion
SKILL.md:15Runtime Dependencies Are Installed Without Version or Integrity Pinning
The activation phrases are very broad ('博客监控 / 网站更新 / 追踪博客') and could trigger the skill in contexts where the user did not clearly intend to invoke external-site monitoring behavior. Because this skill performs network access to user-specified or configured URLs, accidental invocation increases the chance of unintended outbound requests and monitoring actions.
The skill describes monitoring blogs and webpages but does not warn users that it will contact external sites and fetch content from provided URLs. In this context, the code uses requests.get() and feedparser.parse() on supplied URLs, so missing warnings and constraints can lead to unsafe use, including fetching internal, sensitive, or unexpected network resources.
The description is written in Chinese, and the rest of the document presents activation phrases and output format only in Chinese, with no indication that another language can be used. This creates a language/locale constraint without user opt-in or documented justification.