tech-employer-branding
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPERSISTENCE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill requires auditing external, untrusted web surfaces such as anonymous professional forums and employer-review sites. This process creates a vulnerability to indirect prompt injection where malicious instructions could be embedded in the audited data to influence agent behavior. * Ingestion points: Web content from third-party employer-review sites and forums (Step 3). * Boundary markers: No boundary markers or delimiters are instructed for use with untrusted content. * Capability inventory: Capability to write to local files (
devrel-context.md) and call related developer-relations skills. * Sanitization: No sanitization or validation of external content is performed. - [PERSISTENCE]: The 'Memory' instruction directs the agent to persist task and project context in a local file named
devrel-context.md, utilizing file-system write operations to maintain information across sessions.
Audit Metadata