implement

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted external data from GitHub issues (titles, bodies, and comments) using the gh issue view command. This creates an attack surface for indirect prompt injection, where a malicious issue could contain instructions designed to override the agent's behavior.
  • Ingestion points: Data retrieved from GitHub issues in Step 1 of SKILL.md.
  • Boundary markers: The instructions do not define clear boundaries or 'ignore' instructions to isolate untrusted issue content from the system prompt.
  • Capability inventory: The skill has the capability to modify files and execute shell commands (git, gh, rspec, rubocop, slim-lint), which could be abused if an injection is successful.
  • Sanitization: No sanitization or validation of the retrieved issue content is mentioned in the instructions.
  • [COMMAND_EXECUTION]: The skill executes multiple shell commands to interact with the repository (git remote get-url), fetch issue details (gh issue view), and perform implementation verification (rspec, rubocop, slim-lint). While these tools are standard for a development assistant, they provide a powerful interface that executes based on the content of the issue being processed.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 04:36 AM
Security Audit — agent-trust-hub — implement