post-task-verification

Pass

Audited by Gen Agent Trust Hub on Aug 21, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external GitHub resources.\n
  • Ingestion points: Commands such as gh issue view {{ARG1}} and gh pr view {{ARG1}} fetch issue/PR descriptions and comments which are analyzed by the agent.\n
  • Boundary markers: The instructions lack explicit delimiters or warnings to ignore instructions embedded within the fetched external content.\n
  • Capability inventory: The agent has the ability to execute shell commands (git, gh, grep) and project-specific build/test/lint commands.\n
  • Sanitization: No sanitization of the external content is described before processing.\n- [COMMAND_EXECUTION]: The skill frequently uses shell commands to interact with the repository and remote services.\n
  • Evidence: Extensive use of git for branch and history analysis, gh for GitHub integration, and grep for codebase searching.\n
  • Context: The skill also directs the agent to execute project-specific commands for linting, testing, and code generation, which are dynamic based on the project environment.\n- [DYNAMIC_CONTEXT_INJECTION]: The skill utilizes dynamic context injection to gather environment data during the loading phase.\n
  • Evidence: Usage of !git branch --show-current, !git log --oneline -5, and !git diff to populate context headers.\n
  • Context: These commands are used for legitimate repository state awareness and do not incorporate unvalidated user input at load time.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 21, 2026, 02:53 AM
Security Audit — agent-trust-hub — post-task-verification