prompt-templating

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to build prompts by interpolating data from untrusted sources, which creates a surface for indirect prompt injection.\n
  • Ingestion points: The skill gathers data from user messages, template files (e.g., templates/code-review.txt), and external file contents retrieved via the Read tool.\n
  • Boundary markers: The templates do not utilize structural delimiters (such as XML tags) to separate interpolated data from system instructions, making it possible for data to be misinterpreted as commands.\n
  • Capability inventory: The agent using this skill has access to Read, Write, Edit, and Grep tools, which could be misused if a malicious instruction is successfully injected.\n
  • Sanitization: While the SKILL.md file advises the agent to sanitize input, there is no programmatic enforcement of this practice within the skill's static files.\n- [NO_CODE]: The skill consists entirely of markdown documentation, instructional text, and template files. It does not include any executable scripts, binaries, or external package dependencies.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 08:16 AM