The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute a local command-line utility named maestro for tasks such as loading context and searching design databases.
[PROMPT_INJECTION]: The skill processes untrusted external data, including user-provided source code, design specifications, and reference materials during auditing and refinement tasks. This establishes an indirect prompt injection surface where instructions embedded in processed files could potentially influence the agent's behavior.
Ingestion points: The skill reads PRODUCT.md, DESIGN.md, and any user-specified target files or visual specifications.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are used when processing target code.
Capability inventory: The skill has access to Bash, Write, Edit, and request_user_input tools, which could be leveraged if an injection is successful.
Sanitization: There is no evidence of sanitization or filtering of external content before it is processed by the agent.

maestro-impeccable