The Agent Skills Directory

[SAFE]: The skill operates entirely within the local development environment, focusing on generating and managing project specifications and implementation code. All templates and examples are standard engineering artifacts.
[COMMAND_EXECUTION]: The skill includes documentation for checking project state via shell commands. These are standard utility commands (ls, test) restricted to the skill's own local data directory (.ssd/).
[PROMPT_INJECTION]: The skill uses strong instructional language to enforce its methodology (e.g., 'Prime Directive', 'The Law'). These constraints are designed to improve the reliability of the agent's engineering output and do not attempt to override system safety protocols.
[PROMPT_INJECTION]: The skill processes user-provided requirements to generate executable specifications, creating a potential surface for indirect prompt injection.
Ingestion points: User feature descriptions and requirements entering the Schema and Sequence phases.
Boundary markers: The skill employs Zod schemas and Mermaid sequence diagrams as structural boundaries to delineate user-defined logic.
Capability inventory: The agent writes files to the local file system and triggers test/compilation runs as part of the reconciliation loop.
Sanitization: The methodology uses Zod for type-safe data validation and explicit human approval steps between phases to verify the generated specifications.

ssd-engineering