structured-problem-solving

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external materials, including long texts, URLs, and office documents (PDF, Excel, Word), which are then used to generate analysis and prompts for other agents. This creates a surface for indirect prompt injection.\n
  • Ingestion points: references/material-handling.md defines the ingestion of external files and web content into the agent's context.\n
  • Boundary markers: The skill provides instructions to use XML tags like <source_materials> to delimit untrusted content, as seen in references/prompt-handoff.md.\n
  • Capability inventory: The skill has the capability to write multiple files and directories to the local filesystem and invoke the markdown-conversion tool.\n
  • Sanitization: The instructions emphasize manual verification and the separation of facts from opinions but lack programmatic escaping or sanitization of the interpolated data.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 06:04 AM
Security Audit — agent-trust-hub — structured-problem-solving