anchoring
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns such as prompt injection, data exfiltration, or obfuscation were identified across the skill files.
- [EXTERNAL_DOWNLOADS]: The skill references external URLs for documentation and scientific research. These include the vendor's own domain (deciqai.com), official government repositories (sec.gov), and well-known academic/news sources (doi.org, Reuters, Bloomberg, etc.), all of which are considered safe for their intended purpose.
- [NO_CODE]: The skill consists entirely of Markdown files providing instructions and historical examples; no executable scripts, shell commands, or package dependencies are included.
- [PROMPT_INJECTION]: Instructional guidelines for the agent (e.g., 'Coaching Novices' and 'Engine mode') represent standard behavioral steering for educational skills and do not attempt to bypass safety filters or override system constraints.
Audit Metadata