computer-scientist
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill is composed exclusively of Markdown instructions and YAML metadata. It does not contain any scripts, shell commands, or executable code blocks.\n- [SAFE]: No malicious patterns such as prompt injection, obfuscation, or persistence mechanisms were detected. The skill uses structured templates for data handling.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external data from research papers, blogs, and leaderboards. While this constitutes an attack surface for indirect prompt injection, it is the primary and intended function of a research-surveying agent. The risk is mitigated by the skill's focus on high-level orchestration rather than direct execution of untrusted content.\n
- Ingestion points: Phase 1.2 (Sources to Scan) includes external papers and competitor technical reports.\n
- Boundary markers: None explicitly mentioned in the survey phase.\n
- Capability inventory: The skill commissions tasks to Principal Scientist agents.\n
- Sanitization: Not specified; the agent's role is restricted to high-level summarization and prioritization.
Audit Metadata