project-issues-batch

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the parsing of GitHub issue content.
  • Ingestion points: External data enters the agent's context during Phase 1 (gh issue view) and Phase 2 (dependency scanning), where issue titles, bodies, and comments are retrieved from GitHub repositories.
  • Boundary markers: There are no defined delimiters or 'ignore' instructions to prevent the agent from obeying malicious prompts embedded within the issue markdown.
  • Capability inventory: The skill executes Git operations (git checkout, git branch), GitHub CLI commands (gh pr create), and delegates tasks to the issue-to-implementation skill, which likely involves file system writes and code generation.
  • Sanitization: No sanitization or content filtering is implemented for the ingested issue data, allowing an attacker with write access to a GitHub issue to potentially influence the implementation behavior or branch naming via malicious instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 12:40 PM