The Agent Skills Directory

[SAFE]: The skill instructions and associated reference guides (SKILL.md, references/claim-first.md, references/guide.md) are purely informational and focused on creative writing and marketing strategy. No executable scripts, command lines, or subprocess calls were identified.\n- [SAFE]: No patterns of obfuscation, such as Base64 encoding, zero-width characters, or homoglyph attacks, were found in the skill content or metadata.\n- [SAFE]: The skill does not perform any network operations or access sensitive local file paths. It relies on provided text content and research briefs, presenting no risk of data exfiltration or credential harvesting.\n- [SAFE]: While the skill ingests external data (like analytics or repository content) for auditing, it lacks the dangerous capabilities (e.g., shell access, file-writing) required to turn indirect prompt injection into a high-risk vulnerability.\n- [SAFE]: YAML metadata and platform configuration fields are used appropriately to define the skill's purpose and scope without any deceptive or malicious instructions.

landing-copy