The Agent Skills Directory

[SAFE]: The skill performs research tasks using well-known and reputable external services such as Google Patents, Espacenet, Google Scholar, and arXiv. These operations are consistent with the skill's stated purpose of conducting prior art searches.- [DATA_EXPOSURE]: The skill reads invention descriptions from user input or specific local files (INVENTION_BRIEF.md). There is no evidence of unauthorized access to sensitive system files, credentials, or environment variables. Information flows primarily from the user/local context to the research report.- [COMMAND_EXECUTION]: Although the Bash tool is listed as allowed in the frontmatter, the workflow instructions do not include the execution of arbitrary or dangerous shell commands. The usage of tools is restricted to searching, reading, and writing project-specific documentation.- [PROMPT_INJECTION]: The skill ingests data from external websites via WebSearch and WebFetch. While this represents a surface for indirect prompt injection, the risk is mitigated by the use of reputable academic and legal databases as sources. No explicit instructions to bypass safety filters or ignore prior instructions were found.

prior-art-search