external-cannbot-ops-triton-task-extractor

Warn

Audited by Socket on Sep 7, 2026

1 alert found:

Security
SecurityMEDIUM
scripts/validate_task.py

No explicit malicious payload is visible in this module itself (no hardcoded secrets, suspicious domains, or direct exfiltration code). However, it is a high-risk component in a supply-chain context because it performs unsandboxed execution of the validated file via `exec(code, namespace)` and then instantiates/calls user-defined functions and model logic. If an attacker can influence the file being validated, they can achieve arbitrary code execution with the validator’s process privileges. Treat this as a security-critical design requiring strict trust boundaries/sandboxing.

Confidence: 78%Severity: 78%
Audit Metadata
Analyzed At
Sep 7, 2026, 03:55 PM
Package URL
pkg:socket/skills-sh/ascend-ai-coding%2Fawesome-ascend-skills%2Fexternal-cannbot-ops-triton-task-extractor%2F@a4f91acdf18e87d3d51352c6d9df8150049c3905c2109950b41c97427270ce28
Security Audit — socket — external-cannbot-ops-triton-task-extractor