The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user-provided text for the purpose of analysis and rewriting, creating a potential surface for indirect prompt injection.
Ingestion points: User-supplied text is processed by the agent as described in the SKILL.md workflow.
Boundary markers: The instructions do not specify the use of strict boundary markers or delimiters to isolate user text from the agent's internal reasoning or instructions.
Capability inventory: The skill utilizes Bash (to execute the detect.py script), Read, Write, Edit, and AskUserQuestion tools.
Sanitization: The scripts/detect.py utility includes a strip_invisible function that removes Unicode format characters (category 'Cf') and normalizes various non-breaking space characters, which serves as a mitigation against obfuscation and certain injection techniques.

anticopywriting-ai