The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it fetches content from arbitrary external URLs and instructs the agent to use that content to answer questions or process data.
Ingestion points: External content fetched via curl in SKILL.md.
Boundary markers: Absent; there are no instructions to the agent to ignore or delimit instructions found within the fetched content.
Capability inventory: Uses shell_execute with curl for network requests.
Sanitization: Absent; the skill does not specify any validation or filtering of the fetched Markdown content.
[COMMAND_EXECUTION]: The skill utilizes shell_execute to run curl commands with user-provided URLs. While it provides instructions to wrap URLs in quotes to prevent shell meta-character exploitation, this remains a significant execution surface.
[EXTERNAL_DOWNLOADS]: The skill is designed to download data from external services (defuddle.md and r.jina.ai). These are well-known tools for web content extraction, and their use is central to the skill's stated purpose.

web-content-extractor