The Agent Skills Directory

[SAFE]: The skill implements a functional workflow for drafting patent claims using structured local files (e.g., invention disclosures and prior art reports). It processes this data within an established context to produce legal documentation.
[SAFE]: No evidence of data exfiltration, credential harvesting, or malicious command execution was found. All file operations (Read, Write, Edit) and tool calls (mcp__codex__codex) are consistent with the skill's stated purpose of document generation and review.
[SAFE]: While the skill reads external content from invention files, this is inherent to its primary function and does not constitute a high-risk indirect prompt injection surface in this context, as the output is directed to a markdown file and a quality-review model without downstream execution of the contents.

claims-drafting