The Agent Skills Directory

[SAFE]: The skill performs legitimate code quality checks by analyzing source code structure and content. Its instructions are clear, transparent, and aligned with its stated purpose.
[DATA_EXPOSURE]: The skill interacts with local project files, including source code and a configuration file (.codeprobe-config.json). This access is localized to the project environment and does not involve accessing sensitive system credentials (e.g., SSH keys, AWS configs) or personal data.
[COMMAND_EXECUTION]: Although the skill has access to Bash, its instructions are limited to project-based searches and line counting. No dangerous command patterns, shell injection vulnerabilities, or persistence mechanisms were detected.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes external content (the user's code), which is an inherent risk surface for indirect prompt injection. However, the skill lacks high-risk capabilities like network access or system modification, which mitigates potential impact.
Ingestion points: Project source code files and .codeprobe-config.json (found in SKILL.md).
Boundary markers: None explicitly defined in the skill instructions to separate code from instructions.
Capability inventory: Read, Grep, Glob, and Bash tools (found in SKILL.md).
Sanitization: No specific sanitization or filtering of the ingested code content is mentioned.

rightcode-codeprobe-code-smells