docx-processing-anthropic
Audited by Socket on Feb 17, 2026
1 alert found:
Malware[Skill Scanner] Instruction directing agent to run/execute external content All findings: [CRITICAL] command_injection: Instruction directing agent to run/execute external content (CI011) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] [CRITICAL] command_injection: Installation of third-party script detected (SC006) [AITech 9.1.4] The code fragment (being a capability/documentation of a document-processing skill) is benign and coherent with its stated purpose. It outlines legitimate workflows and dependencies for manipulating Word documents, including tracked changes and redlining, without introducing credential requirements or suspicious network activity. LLM verification: No direct malware indicators or obfuscated code are present in the provided SKILL.md text. The skill's functionality (docx processing, tracked-changes, using pandoc and OOXML unpack/pack) is consistent with its stated purpose. However, it instructs installing and executing third-party tooling without specifying trusted sources or integrity checks and mandates reading full reference files without limits; those elements increase software supply-chain risk. Overall the package appears functionally