The Agent Skills Directory

[SAFE]: The skill is purely instructional text designed to guide the agent's reasoning process when analyzing legal documents. It does not perform any autonomous actions.
[NO_CODE]: There are no scripts, binaries, or configuration files that execute code. The analysis is performed entirely within the context of the AI's natural language processing.
[SAFE]: No external network requests or remote data fetching mechanisms are present. The skill relies on data provided directly by the user during the session.
[SAFE]: While the skill is designed to ingest and process untrusted external data (legal documents), it does not have access to any tools (as evidenced by the lack of an allowed-tools field or tool-calling instructions) that could be exploited through indirect prompt injection.

tos-clause-scanner