writing-assistant

Pass

Audited by Gen Agent Trust Hub on Apr 4, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the mcporter Node.js package from the public registry. This is a legitimate utility provided by the vendor to facilitate interaction with their text analysis services.
  • [COMMAND_EXECUTION]: Uses the mcporter CLI to configure access and perform text analysis. These commands are directed to the vendor's official endpoint at mcp.infranodus.com.
  • [PROMPT_INJECTION]: The skill includes specific directives to prevent disclosing its internal logic and instructions to users, mitigating potential system prompt extraction. Additionally, the skill is susceptible to indirect prompt injection as it processes arbitrary user input for text refinement. Ingestion points: Arbitrary text provided by users for grammar and style refinement in SKILL.md. Boundary markers: No specific delimiters are used to isolate user content from the skill's instructions. Capability inventory: Possesses the ability to execute shell commands and call external tools via the mcporter CLI based on text analysis. Sanitization: Does not implement validation or sanitization of input text before processing.
  • [DATA_EXFILTRATION]: User-provided text is sent to the external domain mcp.infranodus.com for analysis. This is a core feature of the skill and uses the vendor's dedicated infrastructure.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 4, 2026, 11:13 PM
Security Audit — agent-trust-hub — writing-assistant