X/Twitter Intelligence Analyst
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data from X/Twitter, creating a surface for indirect prompt injection attacks.
- Ingestion points: Phase 2 (Signal Collection) explicitly instructs the agent to collect posts, threads, and profiles from the live web or external datasets (SKILL.md).
- Boundary markers: The skill lacks explicit instructions for using delimiters or XML-style tags to isolate untrusted data, although it does instruct the agent to "Separate Observation From Interpretation."
- Capability inventory: The skill instructions do not define or call for high-risk capabilities such as arbitrary command execution, local file system writes, or network exfiltration to unverified domains.
- Sanitization: The workflow includes a "Deduplication" and "Cleaning" step to remove spam and irrelevant matches, which provides a basic level of filtering for high-volume noise but is not designed to detect adversarial prompt injection content.
- [SAFE]: The skill mandates security best practices by instructing the agent to protect credentials and use API keys through environment variables or secret stores rather than hardcoding them.
- [SAFE]: The instructions include ethical constraints that forbid harassment, doxxing, and the use of unauthorized data, reducing the risk of the agent being used for malicious purposes.
Audit Metadata