Is hargabyte/chaos-mind safe?

https://github.com/openclaw/skills/tree/main/skills/hargabyte/chaos-mind

87
SAFE

CHAOS Memory is a local memory management system for AI agents that appears well-designed with privacy controls. The main security considerations are the installation of remote binaries and potential access to sensitive session data, but both are handled responsibly with appropriate safeguards and user control.

Category Scores

Prompt Injection 85/100 · 30%
Data Exfiltration 90/100 · 25%
Code Execution 75/100 · 20%
Clone Behavior 95/100 · 10%
Canary Integrity 100/100 · 10%
Behavioral Reasoning 80/100 · 5%

Findings (4)

MEDIUM Remote Binary Download and Execution -25 ▶

The install script downloads pre-built binaries from GitHub releases and executes them locally. While this uses signed releases from a reputable source, it inherently carries risk of supply chain attacks if the release infrastructure were compromised.

LOW Session File Access Capability -10 ▶

The auto-capture feature can read session transcript files containing potentially sensitive conversation data. However, this is disabled by default and requires explicit user configuration of file paths.

LOW Potential Sensitive Data Accumulation -20 ▶

The tool stores memories in a local database and could accumulate sensitive information over time if users store sensitive data or misconfigure auto-capture settings.

INFO Agent Behavior Instructions -15 ▶

The skill provides detailed instructions on how AI agents should use the memory management commands. These appear legitimate and appropriate for the tool's functionality.