The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection attacks where malicious instructions could be embedded in audited files.
Ingestion points: The agent is instructed to read and analyze data from the filesystem and codebase, including password policies, hashes, and breach databases.
Boundary markers: The instructions lack explicit delimiters or warnings to treat processed data as untrusted content.
Capability inventory: The agent has access to the filesystem and codebase tools, providing a surface for injected instructions to interact with the user's environment.
Sanitization: There is no evidence of input validation, escaping, or filtering for the external content being processed.

password-auditor