tech-press-relations
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to browse external websites to verify reporter bylines and employment status. This creates a surface for indirect prompt injection where malicious instructions could be embedded in the external content (e.g., an article byline or an outlet's website). The risk is considered minimal as the capability is necessary for the skill's primary purpose.
- Ingestion points: Web browsing for bylines and employer verification as described in
SKILL.mdandreferences/media-map-template.md. - Boundary markers: Absent.
- Capability inventory: Web browsing, local file writing (
devrel-context.md), and text generation. - Sanitization: Absent.
- [SAFE]: The skill follows professional PR best practices, referencing established industry standards like the PESO model and Barcelona Principles. References to external organizations such as Google, GitHub, GitLab, and Supabase are directed toward official documentation and known technical guides. No code execution, obfuscation, persistence, or sensitive data exfiltration patterns were identified.
Audit Metadata