pp-expensify

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of the expensify-pp-cli and expensify-pp-mcp tools from external sources.
  • Fetches the @mvanhorn/printing-press-library package via npx.
  • Installs Go modules from github.com/mvanhorn/printing-press-library.
  • [REMOTE_CODE_EXECUTION]: The skill executes code downloaded from remote registries during the setup process.
  • npx -y @mvanhorn/printing-press-library install expensify --cli-only downloads and runs the installer from the NPM registry.
  • go install github.com/mvanhorn/printing-press-library/... compiles and installs remote source code from GitHub.
  • [COMMAND_EXECUTION]: The skill is built around executing shell commands via the expensify-pp-cli binary using the Read Bash tool to interact with the Expensify service.
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts natural language prompts and external file data which are passed as arguments to CLI commands.
  • Ingestion points: Arguments for the expense quick and expense from-line commands, and the feedback command's input.
  • Boundary markers: None explicitly implemented to separate instructions from data within the CLI arguments.
  • Capability inventory: The CLI can perform file writes (SQLite database), read local configuration, and execute network requests to the Expensify API and user-defined webhooks.
  • Sanitization: No explicit sanitization or validation of the input strings is mentioned in the skill instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 11:47 AM
Security Audit — agent-trust-hub — pp-expensify