ralph-specum-design

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill instructions require reading external data from files like requirements.md and research.md. This ingestion creates a surface for indirect prompt injection where malicious instructions embedded in those files could influence agent behavior.
  • Ingestion points: Reads requirements.md, research.md, and .progress.md (specified in SKILL.md).
  • Boundary markers: None. The skill does not define delimiters or specific instructions to isolate untrusted data from the system prompt.
  • Capability inventory: The skill is limited to reading and writing markdown files (design.md, .progress.md) and managing internal state. It mentions committing artifacts, which likely involves standard version control tool usage.
  • Sanitization: None. The skill does not specify any validation, filtering, or escaping for the ingested file content.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 07:36 PM
Security Audit — agent-trust-hub — ralph-specum-design