The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill fetches and processes data from the public package registry at packagist.org, which contains user-submitted content that could include malicious instructions.
Ingestion points: Data is ingested through the packagist_org__render_page tool and other scraping functions defined in SKILL.md.
Boundary markers: The instructions do not specify any delimiters or safety warnings for the agent to ignore instructions embedded in the external content.
Capability inventory: The skill allows for searching, scraping, and performing tasks on the remote site, including automated actions through the Unbrowse toolset (SKILL.md).
Sanitization: There is no evidence of sanitization or filtering of the external content before it is passed to the agent.
[EXTERNAL_DOWNLOADS]: The skill instructs the user or agent to connect to a remote Model Context Protocol (MCP) server hosted by the vendor.
The remote server is located at https://unbrowse.ai/mcp/packagist.org and is used to execute the search and interaction tools server-side.

packagist