The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted markdown data from 'brain pages' and uses fragments of that data to perform external searches and rewrite files.
Ingestion points: The skill uses the get_page tool to read markdown content during the 'Scan pages' phase (Phases 1 & 2).
Boundary markers: There are no documented boundary markers or 'ignore' instructions to separate data from commands when processing these pages.
Capability inventory: The skill has file-write capabilities via put_page and network capabilities via the X API integration and search tool.
Sanitization: There is no mention of escaping or sanitizing the content extracted from brain pages before it is used in API queries or written back to the file system.
[COMMAND_EXECUTION]: The skill's documentation includes a shell script snippet for batch processing. While intended for user-initiated bulk cleanup, it demonstrates how the agent can be instructed to perform recursive filesystem scans using find and grep.

citation-fixer