The Agent Skills Directory

[PROMPT_INJECTION]: The skill is designed to process external, untrusted text provided by users for editing. This creates a vulnerability surface for indirect prompt injection, where an attacker could embed malicious instructions within the text meant for humanizing. 1. Ingestion points: Text provided by the user or read from files via allowed tools like Read and Glob. 2. Boundary markers: The instructions lack explicit delimiters or instructions to treat the input as passive data. 3. Capability inventory: The agent is granted powerful file system tools (Read, Write, Edit, Grep, Glob) which could be misused if the agent is subverted. 4. Sanitization: No input sanitization or validation logic is present.
[NO_CODE]: The skill consists entirely of markdown instructions and does not include any scripts, executables, or package dependency files.
[SAFE]: The skill refers to a well-known external source (Wikipedia) for its guidelines. No hardcoded credentials, malicious URLs, or unauthorized network operations were detected.

humanizer