The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a utility module scripts/runtime/common.mjs that defines a runChild function using Node.js spawn. This function is part of the skill's runtime infrastructure for internal checks, but it provides a primitive for executing external programs.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to read, interpret, and summarize data from various project files (such as .devflow/index.md and Sketch handoff finding files) to publish design foundations and decision records. This ingestion of untrusted data constitutes an indirect prompt injection surface.
Ingestion points: Project configuration and artifact files, specifically those found in the .devflow/ directory and referenced in references/sketch-handoff.md.
Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands within the data it processes.
Capability inventory: The skill has capabilities for file system read/write operations and process execution primitives in its utility scripts.
Sanitization: There is no evidence of sanitization or filtering logic applied to the content of the external files before they are processed by the agent.

design