oneshot-gov
Pass
Audited by Gen Agent Trust Hub on Nov 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes solicitation data, such as titles and descriptions, retrieved from external sources. This data is untrusted and could potentially contain malicious instructions designed to influence the agent.
- Ingestion points: Data returned by the
agent.govSolicitationsmethod described inSKILL.md. - Boundary markers: There are no explicit boundary markers or instructions for the agent to ignore embedded commands within the processed solicitation text.
- Capability inventory: The skill documentation suggests integration with the
oneshot-emailskill, which provides the agent with the ability to generate and send emails based on the notice content. - Sanitization: No sanitization or filtering of the solicitation data is mentioned in the skill instructions.
- [EXTERNAL_DOWNLOADS]: The skill integrates the vendor's official SDK for government solicitation searches.
- Evidence: Import of the
@oneshot-agent/sdkpackage in the TypeScript implementation examples.
Audit Metadata