The Agent Skills Directory

[SAFE]: The skill is designed as a pure read-and-report tool. It explicitly states that it writes nothing to disk and only processes local literature notes to identify terms for the user.
[DATA_EXFILTRATION]: No network-enabled tools or functions (such as curl, wget, or fetch) are present. The skill does not attempt to transmit any of the data it reads to external domains.
[COMMAND_EXECUTION]: The skill contains no instructions to execute shell commands, subprocesses, or external scripts.
[PROMPT_INJECTION]: The instructions are clear and purpose-oriented, without any attempts to bypass safety filters, override system prompts, or extract hidden instructions.
[REMOTE_CODE_EXECUTION]: There are no patterns involving the download or execution of remote code, and no external dependencies are specified.

find-terms