planning-jira-tasks

Pass

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it consumes untrusted data from Jira ticket snapshots at docs/<TICKET_KEY>.md. If a ticket contains adversarial instructions in its description or comments, it could influence the subagents' planning output.\n
  • Ingestion points: docs/<TICKET_KEY>.md is read by multiple subagents including task-planner, task-validator, and stage-validator.\n
  • Boundary markers: The instructions lack explicit delimiters or 'ignore embedded instructions' warnings for the ticket content.\n
  • Capability inventory: The skill utilizes file read/write operations via agent tools but does not invoke subprocesses or network operations.\n
  • Sanitization: No sanitization or validation of the text within the ticket snapshot is performed beyond structural presence checks.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 20, 2026, 01:57 PM