security-advisory-review

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill generates and executes local Proof-of-Concept (PoC) scripts inside temporary directories to validate security advisories. It also executes various git commands to manage repositories and worktrees.
  • [EXTERNAL_DOWNLOADS]: The skill clones remote Git repositories from user-provided URLs into the /tmp directory for analysis.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it ingests untrusted data from security advisories and repository source files without defined boundary markers or sanitization logic.
  • Ingestion points: Security advisory text (Phase 1) and repository source files (Phase 3).
  • Boundary markers: None identified; instructions do not specify delimiters to separate untrusted data from system instructions.
  • Capability inventory: Git command execution, file system writes, and shell execution of generated PoC scripts.
  • Sanitization: No explicit sanitization or escaping of external content is mentioned before the agent analyzes or incorporates it into the workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 07:52 AM
Security Audit — agent-trust-hub — security-advisory-review