pp-flipp
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the
flipp-pp-clibinary. It provides instructions to download this tool usingnpxfrom the@mvanhorn/printing-press-librarypackage and viago installfrom thegithub.com/mvanhorn/printing-press-libraryrepository. - [COMMAND_EXECUTION]: All primary functionality is delivered through the execution of the
flipp-pp-cliandflipp-pp-mcpbinaries. This includes searching for deals, managing watchlists, and comparing basket prices. - [DATA_EXFILTRATION]: The CLI includes an
--deliver webhook:<url>feature that allows the agent to POST the output of any command to an external URL. While intended for automation, this provides a mechanism for routing data to remote servers. - [INDIRECT_PROMPT_INJECTION]: The skill ingests unauthenticated data from external Flipp web endpoints (flyers, items, and deals).
- Ingestion points: Data is ingested from Flipp's public web endpoints via the search and flyer list commands (SKILL.md).
- Boundary markers: The skill uses
--agentand--jsonflags to structure output but does not explicitly instruct the agent to ignore instructions that might be embedded in the retrieved grocery data. - Capability inventory: The skill can execute shell commands (
flipp-pp-cli), write files (--deliver file:), and perform network operations (--deliver webhook:and API calls) (SKILL.md). - Sanitization: No explicit sanitization or filtering of the retrieved flyer content is described before it is processed by the agent.
Audit Metadata