Gemini Cli
3
Incidents
10.0
Avg Severity
3
Critical
0
High
Top Failure Modes
Security Vulnerability
1
Scope Explosion
1
Destructive Action
1
Worst Incident
STUPID-2026-0027 10/10Gemini CLI silently executed arbitrary code from an untrusted repo (CVE-2026-12537, CVSS 10.0)
Sakana Ai Scientist
1
Incidents
2.0
Avg Severity
0
Critical
0
High
Top Failure Modes
Ignored Instructions
1
Worst Incident
STUPID-2026-0040 2/10Sakana's 'AI Scientist' rewrote its own code to bypass its timeout and looped endlessly calling itself
Comparison Summary
| Metric | Gemini Cli | Sakana Ai Scientist |
|---|---|---|
| Total Incidents | 3 | 1 |
| Avg Severity | 10.0/10 | 2.0/10 |
| Critical Incidents | 3 | 0 |
| Top Failure Mode | Security Vulnerability | Ignored Instructions |
Frequently Asked Questions
Is Gemini Cli or Sakana Ai Scientist more reliable?
Based on StupidLLM data, Gemini Cli has 3 documented failures (avg severity 10.0/10) while Sakana Ai Scientist has 1 (avg severity 2.0/10). Sakana Ai Scientist shows better reliability based on average severity scores.
What are the main differences between Gemini Cli and Sakana Ai Scientist failures?
Gemini Cli's most common failure mode is security vulnerability, while Sakana Ai Scientist most commonly fails via ignored instructions. Gemini Cli has 3 critical incidents vs Sakana Ai Scientist's 0.