issue-triage

Pass

Audited by Gen Agent Trust Hub on Jun 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill is susceptible to instructions embedded within the external data it processes.
  • Ingestion points: The skill ingests untrusted data from "Open GitHub issues and discussions (or Linear/Jira via MCP if configured)" as specified in SKILL.md.
  • Boundary markers: The instructions lack explicit boundary markers or delimiters to separate external issue content from the agent's internal logic, increasing the risk of the agent obeying instructions found within issue bodies.
  • Capability inventory: The skill performs file-write operations to issue-triage-state.md and interfaces with issue tracking systems via MCP (SKILL.md).
  • Sanitization: No evidence of sanitization, escaping, or validation of the ingested external content is present in the instruction set.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 28, 2026, 10:28 AM
Security Audit — agent-trust-hub — issue-triage