The Agent Skills Directory

[COMMAND_EXECUTION]: The Python script scripts/generate_report_bundle.py performs local file operations to generate report artifacts in Markdown, HTML, and DOCX formats. It does not execute arbitrary shell commands or external scripts based on user input.\n- [DATA_EXFILTRATION]: No network activity, hardcoded credentials, or access to sensitive system paths (e.g., SSH keys or cloud configuration files) were found. The skill operates exclusively on provided local case data.\n- [OBFUSCATION]: The skill instructions and scripts are provided in clear text. No hidden payloads, zero-width characters, or malicious homoglyphs were detected. Binary content in sample report files follows standard format specifications.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied JSON data to generate reports. The report generator properly sanitizes this data using html.escape and lacks exploitable capabilities like network access or shell execution, neutralizing the risk of indirect injection.\n- [DYNAMIC_CONTEXT_INJECTION]: The skill entry file SKILL.md does not contain any shell execution triggers (the !command pattern) that would run at load time.

yao-gametheory-skill