writing-clearly-and-concisely

The skill consists of eight Markdown files: README.md, SKILL.md, and several reference files under elements-of-style/ and signs-of-ai-writing.md. All files are purely informational and instructional text. No executable scripts, binaries, or external code downloads are present. The installation instruction in README.md (cp -r skills/writing-clearly-and-concisely ~/.claude/skills/) is a local file copy operation, which is benign and does not involve external network requests or privilege escalation.

No patterns indicative of prompt injection, data exfiltration, obfuscation, privilege escalation, or persistence mechanisms were found. The signs-of-ai-writing.md file, while discussing various malicious or suspicious patterns (like prompt injection, obfuscation, and external links with tracking parameters), does so in a descriptive context as a guide for detection, not as active instructions for the AI to perform these actions. The external URLs referenced in this guide are to academic papers or news articles, not to executable code or data exfiltration targets.

The skill's functionality relies solely on the AI's ability to process and apply the provided textual instructions and knowledge. As such, it does not introduce any direct security vulnerabilities through its content or operations. The inherent risk of indirect prompt injection, common to any skill that processes user-provided text, is noted as an informational risk of the LLM paradigm itself, but not a specific vulnerability within this skill's design.