regex-vs-llm-structured-text

Pass

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or behaviors detected. The skill outlines a standard architectural framework for text extraction. It follows best practices for efficient LLM usage by delegating deterministic tasks to regex.- [PROMPT_INJECTION]: The skill defines a surface for processing untrusted data in its implementation examples. While it lacks explicit boundary markers or sanitization, there are no exploitable capabilities included in the skill code that would allow an injection to achieve unauthorized actions.
  • Ingestion points: content and original_text parameters in parse_structured_text and validate_with_llm.
  • Boundary markers: Absent in the LLM prompt template within validate_with_llm.
  • Capability inventory: Limited to data transformation and LLM API calls; no OS-level, file system, or network exfiltration capabilities identified across the provided functions.
  • Sanitization: Not implemented in the architectural examples.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 24, 2026, 07:42 AM
Security Audit — agent-trust-hub — regex-vs-llm-structured-text