The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (content extracted from Little Red Book) saved within a local directory.
Ingestion points: The skill reads all .md files located in the ~/Documents/Obsidian Vault/xhs directory to perform analysis and summaries.
Boundary markers: There are no explicit instructions to use delimiters or ignore potentially malicious instructions embedded within the saved bookmarks.
Capability inventory: The skill is granted access to powerful tools including Bash, Read, Glob, and Grep.
Sanitization: No sanitization or filtering is applied to the content before it is passed to the AI for summarization, creating a surface where malicious text in a bookmark could influence the agent's behavior.

xhs-analyze