The Agent Skills Directory

[SAFE]: The skill is primarily instructional, providing detailed methodologies and templates for generating technical bid documents. It does not perform sensitive operations like external network requests, arbitrary code execution, or credential harvesting.
[COMMAND_EXECUTION]: While the skill mentions creating project directories and using OCR tools, these actions are represented as logical workflow steps or suggestions for tool use within the agent's environment, rather than malicious command injection or privilege escalation attempts.
[DATA_EXFILTRATION]: The skill focuses on local document generation. There are no mechanisms detected that would send user-provided tender information or local configuration data to external domains.
[PROMPT_INJECTION]: The instructions are well-defined within the domain of professional bid writing. No adversarial prompt injection techniques, such as instructions to ignore safety filters or leak system prompts, were found.

bidwriter