The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified in the analyzed files. The skill consists entirely of markdown documentation and theoretical frameworks for sustainability analysis.
[PROMPT_INJECTION]: The instructions focus entirely on sustainability workflows and do not attempt to override agent behavior or bypass safety guidelines.
[DATA_EXPOSURE]: No hardcoded credentials, sensitive file paths, or network exfiltration patterns were found. References to paths like .env or .ssh are absent.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote code downloads or execution. All references to other files are relative paths within the local skill environment (e.g., references/sustainability-strategy-framework.md).
[OBFUSCATION]: No obfuscated text, encoded URLs, or hidden characters were detected in any of the files.
[INDIRECT_PROMPT_INJECTION]: While the skill is designed to process external business drafts (Mode B and C), it lacks active capabilities such as file-writing or network operations that could be exploited via injection. It functions primarily as an informational audit tool.

meta-sustainability