The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external data such as activity descriptions and legal documents to generate time entries. While it acts as a data ingestion surface, it lacks dangerous capabilities like network access or shell execution, limiting the impact of potential malicious instructions embedded in the input.
Ingestion points: User-provided activity descriptions and legal documentation referenced in the workflow (SKILL.md).
Boundary markers: No specific delimiters or safety instructions are defined to separate user data from agent instructions.
Capability inventory: Text processing and formatting based on a local template; no network, file-system writing, or command execution tools are utilized.
Sanitization: The instructions do not specify any validation or sanitization for external content.
[NO_CODE]: The skill is purely instructional and does not include any scripts, binaries, or configuration files that execute code.
[SAFE]: No evidence of credential harvesting, unauthorized network communication, or obfuscation was found. The skill adheres to its stated purpose of assisting with legal time tracking.

zeitnarrative