incident-ingesting

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from Jira tickets.
  • Ingestion points: Phase 1 and Phase 2 involve fetching and processing fields like summary, description, and comment from the atlassian MCP server.
  • Boundary markers: Absent. The instructions do not specify the use of delimiters or 'ignore' instructions for the extracted text during processing or handoff.
  • Capability inventory: The skill utilizes the Bash, Read, Write, and Task tools, and facilitates handoffs to other automated components such as the Incident Commander.
  • Sanitization: None. The instructions explicitly require preserving raw error messages and stack traces verbatim, which could harbor malicious instructions.
  • [SAFE]: Connects to the official Atlassian MCP server at https://mcp.atlassian.com/v1/mcp to retrieve issue data.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 04:14 PM