The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool for directory scanning and file path resolution within the .design/ folder. These operations are limited to local project management tasks and do not involve arbitrary command construction from untrusted input.
[DATA_EXFILTRATION]: All data operations are confined to the local filesystem (reading from .design/branding/ and writing to .design/projects/). No network-enabled tools or external URLs are present in the instructions.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from BRIEF.md and config.json which is then passed to a sub-agent (gsp-scoper).
Ingestion points: {PROJECT_PATH}/BRIEF.md and {PROJECT_PATH}/config.json are read and passed to the agent.
Boundary markers: The instructions do not specify explicit delimiters or isolation for the ingested text.
Capability inventory: The skill possesses Bash, Write, and Agent (sub-agent spawning) capabilities.
Sanitization: No explicit sanitization or validation of the brief content is performed before processing. However, the risk is mitigated as the operations are scoped to design document generation within a controlled directory.

project-brief