github-stars-analyzer
Warn
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: MEDIUMPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [Prompt Injection] (MEDIUM): Potential for Indirect Prompt Injection (Category 8). The script
github_api.pyfetches untrusted repository descriptions and metadata from the GitHub API. There are no boundary markers or sanitization logic present to prevent malicious instructions within these external fields from influencing the agent's output or downstream actions. Capability tier is MEDIUM due to the combined ingestion of external data and exfiltration/report generation capability. - [External Downloads] (MEDIUM): The documentation in
SKILL.mdreferences several essential scripts (analyze_repository.py,generate_reports.py,visualize_data.py) that are not included in the provided file set. This obfuscates the core logic of the skill, making it impossible to confirm the absence of dangerous side effects such as insecure file operations or command execution during the analysis or report generation phases.
Audit Metadata