The Agent Skills Directory

[COMMAND_EXECUTION] (SAFE): The validation script validate_brief.sh uses standard shell utilities like grep and wc. It correctly handles input by quoting variables (e.g., "$BRIEF_FILE"), which prevents command injection through malformed filenames.
[REMOTE_CODE_EXECUTION] (SAFE): No remote code execution or package installation patterns were detected. The skill relies on basic system tools available in standard environments.
[DATA_EXFILTRATION] (SAFE): The skill does not perform any network operations. It operates exclusively on the local filesystem to validate the provided markdown template.
[PROMPT_INJECTION] (SAFE): The provided template is for project documentation and contains no instructions designed to manipulate the LLM's system prompt or bypass safety filters.
[INDIRECT_PROMPT_INJECTION] (SAFE): While the script reads external markdown files, it does not interpret their content as instructions. It only matches static patterns to report on documentation quality, presenting a negligible attack surface.

design-brief-generator