The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads and processes the content of an existing spec.md file from a user-specified path. This file content is untrusted and analyzed to generate clarification questions and updates, presenting an attack surface for indirect prompt injection.
Ingestion points: The skill reads spec.md during the 'Initialize' and 'Update Section' steps.
Boundary markers: No delimiters or "ignore embedded instructions" warnings are used when processing the file content.
Capability inventory: The skill utilizes Read, Write, and Bash tools, along with the sequential-thinking MCP tool.
Sanitization: No sanitization, escaping, or validation of the ingested file content is performed.
[COMMAND_EXECUTION]: The skill utilizes a Bash tool for file operations and accepts a feature path via $ARGUMENTS. This path is used to construct the target file path. Without strict validation of the input path, this could potentially be used to target files in unintended directories, although the logic limits the target filename to spec.md within that directory.

clarify