Strategy Workflow

• Analyzed: May 14, 2026, 4:58 AM
• Findings: 1
• Version: 0.1.0
• Hash: de06e7d0c621aea5c5137ce71752190ee116833c4f97caebcf10a0a7ba3aee15
• Source: palm
• External report: View on VirusTotal

Findings (1)

The skill bundle is classified as suspicious due to its extensive use of high-risk capabilities and sophisticated prompt injection techniques, even though the stated purpose is benign. Key indicators include explicit instructions for remote root access and execution via `scp` and `ssh` to `root@HOST` (SKILL.md, backtest-optimize.md), direct commands for data exfiltration to 'mac local machine & cloud' (references/strategy_generation.md), and the 'Autonomous Doctrine (Mandatory)' in backtest-optimize.md. This doctrine is a strong form of prompt injection, instructing the AI agent to operate with mandatory persistence, self-healing, and autonomous decision-making, which represents a significant vulnerability if the agent's directives were to be subverted.