The Agent Skills Directory

[SAFE]: The skill operates entirely within the local project environment, reading provided context files and writing a structured markdown specification to a local directory. It does not utilize network tools, access sensitive system-level credentials, or execute arbitrary code.
[PROMPT_INJECTION]: The skill creates an instruction chain where user input and project files are used to generate a specification that guides a downstream coding agent. While this pattern constitutes an indirect prompt injection surface, it is consistent with the skill's primary function of spec-driven development. 1. Ingestion points: Processes content from project context files (e.g., project-overview.md, architecture.md) and user prompts. 2. Boundary markers: Absent in the generated markdown template, which could lead to instruction confusion for downstream agents. 3. Capability inventory: The skill utilizes file system write capabilities to create new markdown files in the context/specs/ directory. 4. Sanitization: No sanitization or validation of input data is performed before interpolation into the output file.

spec-gen