The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or security vulnerabilities were detected across the skill files.
[EXTERNAL_DOWNLOADS]: The README.md file contains a link to a packaged version of the skill hosted on the author's GitHub repository (github.com/synapseradio/ai-skills). This is documented as a standard installation method for the vendor's resource.
[DATA_EXPOSURE]: The skill is designed to analyze other agent skills within the deployment environment. The instructions in references/audit.md, references/design.md, and references/refactor.md direct the agent to read local files and deployment configurations to perform its core functions of auditing and design research. This behavior is consistent with the skill's stated purpose and occurs within the user's controlled conversation context.
[PROMPT_INJECTION]: The skill contains instructions for evaluating external data (other skills), which technically presents a surface for indirect prompt injection. However, the skill provides a structured methodology (the "ten principles") for the agent to follow, and the risk is inherent to its primary purpose as an auditing tool. No malicious injection patterns or bypass attempts were found.

skill-design