quality-audit

Pass

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local shell commands including unilyze, jq, git, and dotnet to perform static analysis and manage report files.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it reads and processes untrusted project source code.
  • Ingestion points: Source files identified for review in SKILL.md Phase 2.
  • Boundary markers: Absent; no delimiters are used when reading file content.
  • Capability inventory: Includes shell command execution (unilyze, jq, git) and file system access.
  • Sanitization: Absent; the content is analyzed as-is.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 9, 2026, 05:40 PM