The Agent Skills Directory

[SAFE]: No malicious code, scripts, or executable patterns were found in the skill's instructions or methodology files.
[SAFE]: No hardcoded secrets, sensitive path access, or unauthorized exfiltration attempts were detected.
[PROMPT_INJECTION]: The skill processes external financial data (financial filings and web searches), representing a surface for indirect prompt injection. This is effectively mitigated by mandatory evidence-labeling rules, strict data verification protocols, and a compulsory independent review by a secondary agent instance. Ingestion points: Public financial reports and web searches. Boundary markers: Explicit data status labels. Capability inventory: Report generation and information synthesis. Sanitization: Independent review process and primary source verification requirements.

serenity