The Agent Skills Directory

[SAFE]: The skill package contains no executable code, scripts, or binaries. It is composed entirely of Markdown files providing strategic advice for offer construction.- [PROMPT_INJECTION]: The behavior instructions provided (e.g., 'Banned Vocabulary' and 'Core Philosophy') are designed to shape the agent's professional persona and advice quality. There are no attempts to override model safety filters, induce jailbreaks, or extract system prompts.- [DATA_EXPOSURE_AND_EXFILTRATION]: While the skill suggests reading local context files like '.agents/product-marketing.md', these are localized to the project environment for personalization. The skill does not access sensitive system paths, hardcode credentials, or perform network operations.

offers