earnings-analysis-hardened

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. 1. Ingestion points: External financial data is ingested via web searches in Phase 1 and Phase 2 of SKILL.md. 2. Boundary markers: No specific delimiters are defined to separate untrusted external data from the agent's core instructions. 3. Capability inventory: The skill utilizes Python execution for charts and a DOCX generation tool for report delivery. 4. Sanitization: It includes safety guardrails described in SKILL.md and SAFETY.md to resist common attack patterns like data exfiltration and safety bypasses.
  • [COMMAND_EXECUTION]: The skill requires the execution of Python code for data processing and chart generation using libraries like pandas, matplotlib, and seaborn.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 06:31 PM
Security Audit — agent-trust-hub — earnings-analysis-hardened