gdelt-doc-search
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes news metadata and aggregates from GDELT DOC 2.0 via the
tiangong-aiCLI, which creates a potential surface for indirect prompt injection if the ingested metadata contains malicious instructions. - Ingestion points:
tiangong-ai data runoutput as described in SKILL.md. - Boundary markers: The skill provides an extensive "Result boundaries" section that instructs the agent on how to interpret partial results, truncation warnings, and aggregate measures.
- Capability inventory: Retrieval is performed through the
tiangong-aiCLI. - Sanitization: The CLI lints every effective query before network access to ensure valid syntax and parameter usage.
Audit Metadata