Skills
skills/sakanaai/shinkaevolve/shinka-setup/Gen Agent Trust Hub

shinka-setup

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions direct the agent to install the shinka-evolve library from the Python Package Index (PyPI).
  • [COMMAND_EXECUTION]: The workflow involves generating code and executing a local smoke test via python evaluate.py. Evaluator templates also describe the use of subprocess to run candidate programs.
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted user task descriptions to generate code. It uses EVOLVE-BLOCK delimiters to separate code regions. Capabilities include file writing and local execution of the generated scaffolds.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 10:04 AM