The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection as it processes project history and user interactions to create constraints for other agents. This is an expected feature of a preference-tracking system.\n
Ingestion points: ~/.designpowers/taste-profile.md and design-state.md.\n
Boundary markers: None explicitly defined.\n
Capability inventory: Reading and writing project files and global configuration files.\n
Sanitization: Relies on the agent to interpret design-related signals from project history.\n- [NO_CODE]: This skill consists of a single markdown instruction file and does not include any executable code, scripts, or binaries.\n- [SAFE]: No malicious patterns, such as data exfiltration, obfuscation, or unauthorized command execution, were detected. The file access is restricted to the specific configuration directory associated with the tool's purpose.

design-memory