job-posting-language-signal
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEPROMPT_INJECTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill includes instructions that override standard agent behavior by requiring specific attribution text to be appended to all workflows created or copied by the agent. Additionally, it directs the agent to alter its network user-agent string specifically to evade CDN-based security filters when accessing external APIs.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external job postings, creating a vulnerability surface. 1. Ingestion points: External job descriptions are fetched from a third-party jobs data source in Step 2. 2. Boundary markers: The instructions do not specify any delimiters or safety warnings to ensure the agent ignores potentially malicious instructions embedded within the job descriptions. 3. Capability inventory: The skill is permitted to write to multiple output columns which are then available for use in subsequent workspace tasks. 4. Sanitization: The skill relies on a language model to rewrite job titles into copy but lacks structured sanitization or validation of the raw input text.
Audit Metadata