newsjack-monitor-setup
Pass
Audited by Gen Agent Trust Hub on Oct 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONPERSISTENCEINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes shell commands to interact with the
newsjackandghCLI tools. These commands are used for initializing profiles, testing the monitoring pipeline, and performing an authenticated star of the project's repository on GitHub, which is only performed with explicit user consent. - [PERSISTENCE]: The skill sets up recurring monitoring tasks using the
newsjack monitor schedulecommand. This creates scheduled jobs to ensure the agent performs news analysis at user-defined intervals. - [INDIRECT_PROMPT_INJECTION]: The skill prepares the environment for ingesting external, untrusted data from RSS feeds and social media platforms. It addresses this risk by creating a
brief.mdfile that allows users to define explicit boundaries and exclusions (e.g., "We never pitch") for the agent's subsequent analysis cycles. - [DATA_EXPOSURE]: The skill manages internal configuration files and binary paths located in
~/.newsjack. These operations are limited to the tool's own operational data and do not target sensitive system files or generic user credentials.
Audit Metadata