monitor-pipeline

Pass

Audited by Gen Agent Trust Hub on Jun 26, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: Analysis of the skill logic and instructions indicates it performs legitimate monitoring of GitLab pipelines. No evidence of credential exfiltration, unauthorized file system access, or obfuscated malicious code was found.
  • [PROMPT_INJECTION]: The skill ingests GitLab job logs, creating a potential surface for indirect prompt injection. Ingestion points: GitLab job logs (SKILL.md). Boundary markers: None. Capability inventory: GitLab MCP server tools. Sanitization: None. The ingestion of logs is required for the skill's primary purpose of providing failure details to the user.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 26, 2026, 01:25 AM
Security Audit — agent-trust-hub — monitor-pipeline