scientific-validation

Pass

Audited by Gen Agent Trust Hub on Mar 18, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill defines a process for ingesting and validating claims from external sources, which creates a vulnerability surface for indirect prompt injection.
  • Ingestion points: Scientific claims and data are extracted from external documents as described in Phase 1 and 1.5 of SKILL.md.
  • Boundary markers: The skill does not provide instructions for using delimiters or boundary markers to isolate external content from the agent's instructions.
  • Capability inventory: The skill is configured with access to 'Bash', 'Write', and 'Read' tools in the SKILL.md frontmatter.
  • Sanitization: No sanitization or verification steps for the external content are specified.
  • [COMMAND_EXECUTION]: The skill requests the 'Bash' tool and provides Python snippets for statistical analysis. These are high-impact capabilities that, while functional for the skill's purpose, could be exploited if an attacker successfully injects instructions through external documents.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 18, 2026, 11:26 AM