sales-nexter

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a persistent state mechanism using references/learnings.md. The agent is instructed to read this file at the start of each invocation and append new discoveries (tips or gotchas) made during the interaction. This creates an indirect prompt injection surface where untrusted data could be recorded into the persistent context, potentially influencing the agent's behavior in future sessions.
  • Ingestion point: references/learnings.md acts as a long-term context store.
  • Boundary markers: Absent. There are no instructions to treat appended content as untrusted or to ignore instructions embedded within the recorded learnings.
  • Capability inventory: The skill facilitates interactions with WordPress REST APIs and configuration of MCP server tools.
  • Sanitization: Absent. The agent is not directed to validate or filter content before appending it to the persistent learning file.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 04:16 PM
Security Audit — agent-trust-hub — sales-nexter