The Agent Skills Directory

[SAFE]: No malicious patterns, executable code, or suspicious commands were detected in the skill files.\n- [NO_CODE]: The skill consists entirely of natural language instructions and YAML configuration. No scripts or binaries are included.\n- [INDIRECT_PROMPT_INJECTION]: The skill is instructed to read from a local file .agents/marketing-context.md if it exists. This represents a potential surface for indirect prompt injection if that file contains malicious instructions. However, because the skill does not have any dangerous tools enabled (such as shell execution or network access), the risk is negligible.

launch-pr