fact-verification
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user drafts and content from arbitrary URLs via the webfetch tool, creating a surface for potential injection attacks.
- Ingestion points: User-provided draft text and external content retrieved via webfetch as described in SKILL.md.
- Boundary markers: Absent; there are no specific instructions to delimit the draft content or to ignore embedded instructions within it.
- Capability inventory: The skill has the ability to fetch external network data, write files to the local system (FACT-CHECK-*.md), and execute shell commands (python scripts/report_publish.py).
- Sanitization: Absent; the skill does not specify any sanitization or validation of the draft content before processing it or passing it to subsequent scripts.
- [COMMAND_EXECUTION]: The skill explicitly instructs the agent to execute a shell command (python scripts/report_publish.py .md) to finalize the report generation process.
Audit Metadata