skills/legnext-ai/skills/legnext/Gen Agent Trust Hub

legnext

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely documentation-based and does not include any executable scripts or binary files.
  • [DATA_EXFILTRATION]: The skill references external endpoints (api.legnext.ai, docs.legnext.ai) and documentation resources. All referenced domains are verified vendor-owned infrastructure for the Legnext service.
  • [CREDENTIALS_UNSAFE]: The documentation promotes secure credential management, explicitly instructing users to use environment variables (LEGNEXT_API_KEY) and avoid committing keys to source control or logs.
  • [PROMPT_INJECTION]: While the skill describes a pipeline for processing user-generated prompts (an indirect prompt injection surface), it provides proactive mitigation guidance. The production-checklist.md recommends treating all user-supplied input as untrusted data and specifically warns against using eval() or creating SSRF vulnerabilities.
  • [SAFE]: The skill provides clear guidance on identifying and handling content moderation errors from the upstream provider, categorizing them correctly as user errors rather than service failures.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 07:25 AM
Security Audit — agent-trust-hub — legnext