pp-policy-intel
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the user or agent to install the
policy-intel-pp-clitool usingnpxfrom the@mvanhorn/printing-press-libraryregistry or viago installfrom the vendor's GitHub repository (github.com/mvanhorn/printing-press-library). These are verified as vendor-controlled resources. - [COMMAND_EXECUTION]: The skill defines several recipes that execute shell commands via the
policy-intel-pp-clibinary. These commands take user-supplied arguments such as search terms, docket IDs, and dates to query federal databases. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources, specifically public rulemaking data from FederalRegister.gov and Regulations.gov.
- Ingestion points: Data is returned to the agent context via the output of
policy-intel-pp-cli(SKILL.md). - Boundary markers: None explicitly defined in the recipes.
- Capability inventory: The agent has access to the
Bashtool (SKILL.md). - Sanitization: The CLI uses an
--agentflag for compact JSON output, but sanitization of document content for prompt injection is not specified.
Audit Metadata