prospero-types-skill

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: Exposure of developer personal information and system configuration in log files.\n
  • The file firebase-debug.log contains internal authorization logs that expose a developer's email address (generamics@gmail.com).\n
  • The log also details requested OAuth scopes and internal API interactions, which provides unnecessary insight into the development environment and internal system behavior.\n- [PROMPT_INJECTION]: Vulnerability to instructions embedded in processed source data (Indirect Prompt Injection).\n
  • Ingestion points: The skill instructions in SKILL.md direct the agent to use read_file and search_file_content to retrieve information from TypeScript files throughout the src/ directory.\n
  • Boundary markers: Absent. The skill does not implement delimiters or specific safety instructions to prevent the agent from obeying instructions that could be hidden within code comments or strings of the analyzed files.\n
  • Capability inventory: The agent is empowered to read local source files and summarize their contents for the user.\n
  • Sanitization: No sanitization or filtering is performed on the ingested source code content before it is interpreted by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 02:10 AM