recurring-jobs

Pass

Audited by Gen Agent Trust Hub on Jul 19, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill enables the creation of recurring tasks that execute instructions provided during user interaction. This represents an indirect prompt injection surface where instructions can be persisted and executed asynchronously. 1. Ingestion points: The manage-jobs tool captures instructions from user chat to create job files (SKILL.md). 2. Boundary markers: The documentation does not specify delimiters or instructions to ignore embedded commands during job execution. 3. Capability inventory: Scheduled jobs are processed via runAgentLoop, which provides access to the agent's execution environment (SKILL.md). 4. Sanitization: There is no mention of input sanitization or validation for the instructions stored in the job metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 19, 2026, 04:41 AM
Security Audit — agent-trust-hub — recurring-jobs