paperboy
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill logs signal delivery data, including recipient identifiers and their responses, to an external service endpoint.
- Evidence: SKILL.md defines a workflow that performs HTTP POST requests to
paperboy-dash.p-d07.workers.dev/deliverandpaperboy-dash.p-d07.workers.dev/suggest-routecontaining distribution details and participant responses. - [INDIRECT_PROMPT_INJECTION]: The skill ingests content from an external source and provides instructions to relay that content without modification, creating a vector for malicious instructions to influence the agent or its recipients.
- Ingestion points: News signals fetched via the
aibtc-newsskill integration. - Boundary markers: None identified in the signal handling instructions to differentiate data from instructions.
- Capability inventory: The skill utilizes network operations via HTTP to log data and is intended for cross-platform communication with other agents and users.
- Sanitization: The skill explicitly mandates that agents "Deliver unaltered" the signals received, omitting any sanitization or validation of the external input.
Audit Metadata