llms-txt

Pass

Audited by Gen Agent Trust Hub on May 13, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface because it processes untrusted data from project documentation and live website content.
  • Ingestion points: SKILL.md (Phases 1, 2, and 4) specifies reading local files like README.md and fetching content from live URLs.
  • Boundary markers: No specific boundary markers or instructions to ignore embedded instructions are implemented when processing these data sources.
  • Capability inventory: The skill uses file-read, file-write (Phase 5), and network-fetch (Phases 2 and 4) capabilities.
  • Sanitization: No explicit sanitization or filtering of external content is mentioned before it is utilized by the agent.
  • [DATA_EXFILTRATION]: The skill facilitates the movement of local project data to public directories (e.g., /public/llms.txt). This presents a risk of data exposure if sensitive information from internal documentation or configuration files is inadvertently identified as project context and published.
Audit Metadata
Risk Level
SAFE
Analyzed
May 13, 2026, 07:26 PM