zafer-skills
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to interpolate a user-supplied project name (
app-name) and bundle ID directly into shell commands and configuration files. - Ingestion points: User input for the project name and bundle ID requested in
SKILL.md. - Boundary markers: No specific delimiters or safety warnings are provided to prevent a user from supplying malicious command arguments instead of a project name.
- Capability inventory: The skill utilizes
bunx,npx,grep, andrmacross its setup and maintenance instructions. - Sanitization: No explicit instructions are given for the agent to validate or escape these inputs before execution.
- [COMMAND_EXECUTION]: The skill relies on multiple shell commands to manage the mobile development workflow.
- Instructions include project initialization (
bunx create-expo), file cleanup (rm src/app/index.tsx), and environment synchronization (npx expo prebuild --clean). - While these are standard operations for Expo development, they grant the agent significant control over the local file system and environment.
- [EXTERNAL_DOWNLOADS]: The skill fetches a sample video from Google's public media storage to implement the required onboarding screen.
- Source:
https://commondatastorage.googleapis.com/gtv-videos-bucket/sample/BigBuckBunny.mp4. - This reference targets a well-known service for sample assets.
Audit Metadata