The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted data from the user's codebase to inform its design and implementation phases. This constitutes a standard functional requirement for a developer tool but establishes a theoretical attack surface.
Ingestion points: The skill reads CLAUDE.md, project README files, UI views, data models, and platform-specific manifest files like Info.plist and AndroidManifest.xml in Phase 1.
Boundary markers: There are no explicit instructions or delimiters provided to the agent to treat the content of these files as untrusted or to ignore instructions contained within them.
Capability inventory: The agent possesses capabilities to create new source code files, modify existing navigation logic, and persist state within the platform's memory system.
Sanitization: The instructions do not specify any validation or sanitization steps for the data extracted from the codebase before it is used in the prompt construction for subsequent phases.
[SAFE]: All external URL references (branding guidelines and design assets) target the vendor's own repository (github.com/iblai/vibe), which is consistent with the skill's authorship and intended purpose. The referenced Node.js packages are either standard developer tools or official vendor SDKs.

iblai-vibe-onboard