The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because its core function involves ingesting untrusted data from the public internet.
Ingestion points: The summarize.py script uses the requests library to fetch content from user-provided URLs.
Boundary markers: The skill does not implement delimiters or 'ignore instructions' warnings when presenting the fetched content to the agent, which could allow a malicious webpage to influence the agent's behavior.
Capability inventory: The skill has network access and local file system write permissions for its caching mechanism located in the user's home directory.
Sanitization: There is no sanitization or instruction filtering performed on the retrieved text before it is summarized and returned to the agent.

summarize-skill