article-figure-spot

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMDATA_EXFILTRATIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • DATA_EXFILTRATION (MEDIUM): The skill is instructed to load credentials from .env files in both the current working directory and the skill's directory. Accessing sensitive files containing API keys is a high-risk activity, though it is tied to the skill's primary function of calling the OpenAI API.
  • COMMAND_EXECUTION (LOW): The skill relies on executing a local Node.js script (scripts/generate-image.mjs) to perform image generation. This is an intended capability but grants the agent command execution power over the local environment.
  • PROMPT_INJECTION (LOW): The skill is susceptible to indirect prompt injection because it processes untrusted article content to generate downstream prompts and filenames.
  • Ingestion points: Article content loaded from file paths or pasted text in SKILL.md (Step 1).
  • Boundary markers: Absent; there are no specific instructions to delimit or ignore instructions within the source text.
  • Capability inventory: Execution of a Node.js script and file system write operations across all provided files.
  • Sanitization: Absent; the skill does not specify any validation or sanitization of the article content before using it to generate prompts or filenames.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:35 PM