skills/zircote/mnemonic/qmd-setup/Gen Agent Trust Hub

qmd-setup

Warn

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the @tobilu/qmd package globally from the NPM registry.
  • [EXTERNAL_DOWNLOADS]: The setup process involves downloading approximately 2 GB of GGUF models for generating embeddings.
  • [COMMAND_EXECUTION]: The instructions direct the agent to execute bash tools/qmd-setup.sh. This file is not provided within the skill's source files, making its behavior and safety unverifiable.
  • [PROMPT_INJECTION]: The skill processes 'mnemonic memories' from various file system paths (e.g., ~/.local/share/mnemonic, .claude/mnemonic/). Since this external data is ingested without boundary markers or explicit sanitization, it presents a surface for indirect prompt injection where instructions embedded in memories could influence agent behavior.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 13, 2026, 03:47 PM
Security Audit — agent-trust-hub — qmd-setup