writing-press-releases
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides links to well-known PR resources and services (AP Stylebook, PR Newswire, Business Wire, HARO) for reference, which is standard practice for this use case.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided information (5 W's, company facts) to generate press releases. While this presents a surface for processing untrusted data, the skill has no dangerous capabilities (e.g., file writing, network requests, or command execution) that could be exploited through the generated text.
- Ingestion points: User input provided during the "Gather Key Information" step (SKILL.md).
- Boundary markers: The skill uses structured Markdown templates to separate user data from generated content.
- Capability inventory: No subprocess calls, network operations, or file system modifications are present.
- Sanitization: None required as the output is static text intended for the user.
Audit Metadata