ai-development-guide

Pass

Audited by Gen Agent Trust Hub on Jul 22, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill instructions create an indirect prompt injection surface by requiring the agent to ingest and act upon content from the repository being modified.
  • Ingestion points: The agent reads source code files (.ts, .tsx), package.json manifests, and CI configuration files to determine quality tool settings.
  • Boundary markers: No explicit delimiters or instructions to ignore embedded instructions are defined for processing external repository data.
  • Capability inventory: The agent is instructed to use shell commands for code searching (grep) and to execute potentially arbitrary scripts defined in package.json.
  • Sanitization: The skill lacks requirements for sanitizing or validating repository content or script definitions before they are processed or executed.
  • [COMMAND_EXECUTION]: The frontend reference guide directs the agent to execute shell-based searches and run project scripts (lint, build, test) using the project's package manager, which involves executing code defined in local configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 22, 2026, 10:14 PM
Security Audit — agent-trust-hub — ai-development-guide