The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains instructions to execute a Python script (scripts/copy_scan.py) that is not included in the provided file set.
Evidence: Found in SKILL.md and references/copy-tells.md as python3 scripts/copy_scan.py <file>.
Context: While the allowed-tools configuration restricts shell-access tools, providing instructions to execute non-existent or external scripts presents a risk if environmental controls are bypassed.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it is designed to ingest and rewrite untrusted user-supplied text.
Ingestion points: Untrusted text is ingested via the 'When given text to humanize' task in SKILL.md.
Boundary markers: No explicit delimiters or boundary markers are defined to isolate the user text from the agent's instructions.
Capability inventory: The agent has access to Read, Write, and Edit tools, which could be leveraged to perform unauthorized file operations if instructions within the input text are followed.
Sanitization: There is no evidence of sanitization or input validation to mitigate malicious instructions embedded in the text being processed.

humanizer