pp-mobbin
Warn
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: MEDIUMCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill primarily functions by executing the
mobbin-pp-clitool to interact with the Mobbin service. - [EXTERNAL_DOWNLOADS]: Instructions are provided to install dependencies from the vendor's repository via
npxandgo installfromgithub.com/mvanhorn/printing-press-library. These resources are provided by the author of the skill. - [CREDENTIALS_UNSAFE]: The authentication mechanism (
auth login --chrome) retrieves sensitive session cookies from the user's Chrome browser profiles to authenticate requests to the Mobbin service. - [DATA_EXFILTRATION]: The CLI includes a
--deliver webhook:<url>flag that allows command output to be automatically POSTed to a user-specified network endpoint. - [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it ingests untrusted UI metadata and OCR text from the Mobbin library.
- Ingestion points: UI screens, flows, and metadata retrieved via
mobbin-pp-clicommands. - Boundary markers: The skill instructions do not define delimiters to separate retrieved tool output from the agent's context.
- Capability inventory: The agent has access to shell commands, file system writes, and network webhooks.
- Sanitization: No content filtering or validation of the retrieved Mobbin content is described.
Audit Metadata