The Agent Skills Directory

[SAFE]: The skill provides a structured workflow for project discovery and initialization. It guides the agent to ask clarifying questions about product scope and developer expertise before making recommendations.
[INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface for indirect prompt injection as it processes untrusted user input during the discovery phase and persists derived information into project files (e.g., .ai/project.json, DECISIONS.md). However, this is inherent to the skill's purpose as a project bootstrap assistant and follows standard documentation practices.
Ingestion points: User responses to product discovery and expertise assessment questions.
Boundary markers: The skill does not define specific delimiters or instructions to ignore potential injection within user responses.
Capability inventory: The agent is instructed to perform file-write operations to create memory and architecture files in the project root and .ai/ directory.
Sanitization: No explicit sanitization or validation of user-provided content is defined.

project-init