The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted user-provided 'project briefs' to generate structured output.
Ingestion points: Project briefs provided by the user are used as the primary input for transformation into specifications.
Boundary markers: The skill does not define explicit delimiters or instructions to the LLM to ignore embedded commands within the input brief.
Capability inventory: The skill has the capability to write to the local file system (docs/specification.md) and automatically trigger subsequent skills (Skill(attune:project-planning)).
Sanitization: No evidence of sanitization or validation of the input content before processing or interpolation into prompts.
[COMMAND_EXECUTION]: The skill documentation explicitly instructs the agent or user to perform local command execution for verification.
Evidence: Several sections contain 'Verification' steps suggesting execution of commands with the --help flag or running pytest -v to confirm success.

project-specification