notepad-system

Pass

Audited by Gen Agent Trust Hub on Apr 24, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a context restoration protocol that reads data from a local file and instructs the agent to treat it as high-priority context.
  • Ingestion points: The agent reads state information from ~/.claude/projects/<project-hash>/notepad.md during the recovery phase.
  • Boundary markers: The protocol uses Markdown headers (# Notepad, ## Priority Notes) to organize data, but it lacks explicit protective delimiters or instructions to ignore potential commands embedded within the notes.
  • Capability inventory: The skill performs file system read/write operations and directly interpolates retrieved content into the active agent context.
  • Sanitization: There is no validation or sanitization of the content retrieved from the notepad file, as the goal is to faithfully restore previous session state.
  • [COMMAND_EXECUTION]: The skill documentation provides shell command examples for manually interacting with the notepad storage.
  • Evidence: Use of the cat command is referenced for reading the state file at ~/.claude/projects/<hash>/notepad.md.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 24, 2026, 05:44 PM
Security Audit — agent-trust-hub — notepad-system