The Agent Skills Directory

Data Ingestion Surface (Indirect Prompt Injection): The skill processes brand guidelines from local files and session history to influence content generation. This mechanism allows external data to guide the agent's output, which is a potential surface for indirect prompt injection. Ingestion points: Accesses .claude/brand-voice-guidelines.md, .claude/brand-voice.local.md, and session context as described in SKILL.md. Boundary markers: No specific delimiters or instructions are used to isolate the ingested guidelines from the agent's core logic. Capability inventory: The skill is restricted to text generation for emails, posts, and proposals, with no access to network resources or code execution tools. Sanitization: No validation or filtering is performed on the guideline content before it is used by the agent.

brand-voice-enforcement