export
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acts as a data bridge for research content, creating a surface where malicious instructions from external sources could be passed through to other skills.\n
- Ingestion points: Research data (including external URLs and summaries) enters the agent context from the filesystem via the
SessionManager.load_latest()method in SKILL.md.\n - Boundary markers: The skill structures output using YAML frontmatter but lacks security delimiters or explicit instructions for downstream agents to ignore potential instructions embedded in the findings.\n
- Capability inventory: The skill possesses the capability to write files to the local filesystem using
pathlib.Path.write_text.\n - Sanitization: There is no evidence of content sanitization or escaping of the findings before they are persisted to the export file.
Audit Metadata