quick-healthy-recipes

Fail

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: HIGHREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The SKILL.md file provides installation instructions that use a piped shell command to execute a script directly from a remote server: curl -fsSL https://get.lightpanda.io | sh. This pattern is used to install the lightpanda CLI but executes unverified code from an external domain with the user's shell privileges.
  • [COMMAND_EXECUTION]: The skill's core workflow relies on executing the lightpanda command-line tool via the shell (e.g., lightpanda fetch --dump markdown <url>) to retrieve external web content for processing.
  • [EXTERNAL_DOWNLOADS]: The skill fetches data and potentially images from external recipe websites and third-party URLs. While specific instructions are provided to respect licensing, the action involves connecting to untrusted remote sources.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes markdown data from external websites retrieved via the lightpanda tool.
  • Ingestion points: External recipe URLs and web content fetched at runtime (documented in SKILL.md).
  • Boundary markers: None identified; external content is parsed directly for recipe extraction without explicit delimiters or instructions to ignore embedded commands.
  • Capability inventory: The agent utilizes network fetching and shell command execution via the lightpanda tool.
  • Sanitization: No validation or sanitization of the markdown content fetched from remote sources is specified in the instructions.
Recommendations
  • HIGH: Downloads and executes remote code from: https://get.lightpanda.io - DO NOT USE without thorough review
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 23, 2026, 10:26 AM
Security Audit — agent-trust-hub — quick-healthy-recipes