The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions do not contain any patterns attempting to override safety guidelines or bypass agent constraints. It uses standard instructional language to define research methodologies and output formats.
[DATA_EXFILTRATION]: There are no commands or instructions that access sensitive files or perform network operations. The skill focuses on processing user-provided research notes locally.
[REMOTE_CODE_EXECUTION]: No external scripts or packages are downloaded or executed. The skill consists entirely of Markdown and JSON configuration files.
[OBFUSCATION]: The content is clear and uses standard Markdown formatting. No hidden characters, Base64 encoding, or homoglyph attacks were detected.
[COMMAND_EXECUTION]: The skill does not instruct the agent to execute any shell commands or interact with the operating system beyond standard text processing.
[INDIRECT_PROMPT_INJECTION]: While the skill is designed to process untrusted user data (such as field notes), it lacks the necessary capabilities (network access, file writing, or command execution) to be exploited through this vector. The instructions include explicit guidelines to distinguish between observed data and interpretation, which serves as a natural boundary for processing external content.

ethnographer