construir
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the terminal to prepare environments, install dependencies, and run the student's project code (e.g., executing
java Main.javaornpxcommands). This is a core feature of the skill intended to make programming concepts observable through execution. - [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes code written by students and existing repository content. While this creates an attack surface for instructions embedded in data, the skill uses structured 'Guided Code' patterns (comments and scaffolds) and a checkpoint-based workflow to maintain control over the agent's logic. Evidence includes ingestion from project files and student implementation tasks, with capabilities involving terminal execution and file writing.
- [DATA_EXPOSURE]: The skill maintains a persistent state in a specific directory (~/.kipperacademy/aprender) to track the student's progress, missions, and learning records. This data is used to provide continuity across sessions and is scoped to the skill's own operational requirements.
- [EXTERNAL_DOWNLOADS]: The skill references and downloads pedagogical resources from the 'kipperacademy' organization. These downloads are performed via established tools like npx and are sourced from the verified author's infrastructure.
Audit Metadata