mastermind-logger

Pass

Audited by Gen Agent Trust Hub on Apr 23, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to create directories and write timestamped markdown files to the local file system under the docs/knowledgelib/ path. It explicitly instructs the agent to skip user permission for these operations ("Do not ask for permission"), which reduces user oversight of file system modifications.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8). It ingests conversation history to generate a knowledge base intended to influence future agents.
  • Ingestion points: The agent scans conversation history and the current file context for decisions, constraints, and learnings.
  • Boundary markers: The skill uses markdown structural headers (e.g., "## Key Decisions") but does not include explicit instructions for future agents to ignore or sanitize embedded commands within the logs.
  • Capability inventory: The skill possesses the capability to write files to the disk autonomously.
  • Sanitization: There is no evidence of sanitization, escaping, or validation logic to ensure that malicious instructions within the session history are not persisted as authoritative "learnings" for future sessions.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 23, 2026, 09:39 PM
Security Audit — agent-trust-hub — mastermind-logger