generate-custom-instructions-from-codebase

Pass

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill is composed entirely of natural language instructions and prompt templates. No executable code, scripts, or automated shell commands are included.
  • [NO_CODE]: There are no functional code components (Python, JavaScript, Shell scripts) provided with this skill.
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes project source code to identify transformation patterns, creating an attack surface for indirect prompt injection from codebase content.
  • Ingestion points: Project codebase evolution between ${SOURCE_REFERENCE} and ${TARGET_REFERENCE} (SKILL.md).
  • Boundary markers: None explicitly defined to separate untrusted code content from the agent's instructions.
  • Capability inventory: The skill requires read access for codebase analysis and write access to create the .github/copilot-migration-instructions.md file.
  • Sanitization: No sanitization or filtering of code comments or content is performed.
  • Context: The risk is assessed as safe because the generated output is a markdown documentation file rather than executable code, and it relies on standard AI documentation workflows.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 18, 2026, 04:15 AM
Security Audit — agent-trust-hub — generate-custom-instructions-from-codebase