the-plant-journal

Pass

Audited by Gen Agent Trust Hub on Sep 24, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest content from local files and live web searches, which constitutes an attack surface for indirect prompt injection if those sources contain adversarial instructions.
  • Ingestion points: The skill (SKILL.md) references external files at ../../resources/source-basis.md and ../../resources/official-source-map.md, and directs the agent to search live sites for author guidelines.
  • Boundary markers: Absent. The instructions do not define delimiters or provide 'ignore embedded instructions' warnings for the data being read.
  • Capability inventory: The skill describes editorial logic and does not directly execute subprocesses, network requests, or file-write operations itself, though it assumes the agent has search and read capabilities.
  • Sanitization: Absent. There are no instructions to validate, filter, or escape the content retrieved from external files or web searches before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 24, 2026, 11:29 AM
Security Audit — agent-trust-hub — the-plant-journal